Skip to content

Latest commit

 

History

History
1808 lines (1506 loc) · 118 KB

alerts.md

File metadata and controls

1808 lines (1506 loc) · 118 KB

Explanations and examples of v-annotate.pl detailed alert and error messages


Output files with detailed alert and error messages

v-annotate.pl outputs two types of files with detailed alert/error messages:

  • .alt files
  • .alt.list files
  • GenBank submission portal detailed error report .tsv files

Documentation on the format of .alt files can be found here and for .alt.list files can be found here. The GenBank submission portal detailed error report .tsv files are the same format as the .alt.list files.

This page includes examples of many of the different alerts and corresponding .alt file output below.

Explanation of sequence and model coordinate fields in .alt files

alert code(s) alert desc(s) sequence coords description model coords explanation link to example
fsthicft, fsthicfi, fstlocft, fstlocfi, fstukcft, fstukcfi POSSIBLE_FRAMESHIFT_HIGH_CONF, POSSIBLE_FRAMESHIFT_LOW_CONF, POSSIBLE_FRAMESHIFT sequence positions of the frameshifted region model (reference) positions of the frameshifted region, some nucleotides may be inserted before or after these positions frameshift example
insertnn, insertnp INSERTION_OF_NT sequence positions of inserted nucleotides with respect to the model model (reference) position after which insertion occurs (always length 1) large insertion example
deletinn, deletinp DELETION_OF_NT sequence position just prior to (5' of) deletion with respect to the model (always length 1) model (reference) positions that are deleted in sequence large deletion example
mutstart MUTATION_AT_START sequence positions of predicted start codon (length <= 3) model (reference) positions that align to the predicted start codon mutated start codon example
mutendcd MUTATION_AT_END sequence positions of predicted stop codon (length <= 3) model (reference) positions that align to the predicted stop codon stop codon alert examples
mutendex MUTATION_AT_END sequence positions of 5'-most in-frame stop codon in the CDS, this stop codon will be 3' of expected stop codon position (always length 3) model (reference) positions that align to stop codon in sequence coords stop codon alert examples
mutendns MUTATION_AT_END will be blank (-) will be blank (-) stop codon alert examples
unexleng UNEXPECTED_LENGTH sequence positions of the predicted CDS, the length of which is not a multiple of 3 model (reference) positions that the predicted CDS align to, some nucleotides may be inserted before or after these positions stop codon alert examples
cdsstopn CDS_HAS_STOP_CODON sequence positions of the 5'-most in-frame stop codon in the CDS, this stop will be 5' of expected stop codong position (always length 3) model (reference) positions that align to stop codon in sequence coords stop codon alert examples
indf5gap INDEFINITE_ANNOTATION_START sequence position of first nucleotide aligned 3' of gap that aligns to the feature boundary (always length 1) model (reference) position of the 5' feature boundary (always length 1) stop codon alert examples
indf5lcc INDEFINITE_ANNOTATION_START sequence position of nucleotide aligned at the 5' feature boundary (always length 1) model (reference) position of the 5' feature boundary (always length 1) examples of indefinite annotation at start
indf5pst INDEFINITE_ANNOTATION_START sequence positions of the nucleotide alignment of the 5' end of the CDS not covered by the protein-based alignment model (reference) position of the 5' boundary of the CDS (always length 1) examples of indefinite annotation at start
indf5plg INDEFINITE_ANNOTATION_START sequence positions of the protein-based alignment not covered by the nucleotide alignment at the 5' end of the CDS model (reference) position of the 5' boundary of the CDS (always length 1) examples of indefinite annotation at start
indf3gap INDEFINITE_ANNOTATION_END sequence position of final nucleotide aligned 5' of gap that aligns to the feature boundary (always length 1) model (reference) position of the 3' feature boundary (always length 1) examples of indefinite annotation at end
indf3lcc INDEFINITE_ANNOTATION_START sequence position of nucleotide aligned at the 3' feature boundary (always length 1) model (reference) position of the 3' feature boundary (always length 1) examples of indefinite annotation at end
indf3pst INDEFINITE_ANNOTATION_START sequence positions of the nucleotide alignment of the 3' end of the CDS not covered by the protein-based alignment model (reference) position of the 3' boundary of the CDS (always length 1) examples of indefinite annotation at end
indf3plg INDEFINITE_ANNOTATION_START sequence positions of the protein-based alignment not covered by the nucleotide alignment at the 3' end of the CDS model (reference) position of the 3' boundary of the CDS (always length 1) examples of indefinite annotation at end
ambgnt5f, ambgnt5c AMBIGUITY_AT_FEATURE_START, AMBIGUITY_AT_CDS_START sequence position(s) of stretch of 1 or more consecutive ambiguous (non-ACGTU) nts ending at the 3' end of a feature model (reference) position(s) the sequence position(s) in sequence coords are aligned to examples of ambiguous nucleotides at start
ambgnt5s AMBIGUITY_AT_END sequence position(s) of 1 or more consecutive ambiguous (non-ACGTU) nts starting at position 1 of the sequence model (reference) position(s) the sequence position(s) in sequence coords are aligned to examples of ambiguous nucleotides at start
ambgnt3f, ambgnt3c AMBIGUITY_AT_FEATURE_END, AMBIGUITY_AT_CDS_END sequence position(s) of 1 or more consecutive ambiguous (non-ACGTU) nts starting at the predicted 5' end of a feature model (reference) position(s) the sequence position(s) in sequence coords are aligned to examples of ambiguous nucleotides at end
ambgnt3s AMBIGUITY_AT_END sequence position(s) of 1 or more consecutive ambiguous (non-ACGTU) nts ending at the final position of the sequence model (reference) position(s) the sequence position(s) in sequence coords are aligned to examples of ambiguous nucleotides at end
ambgcd5c AMBIGUITY_IN_START_CODON sequence position(s) of start codon model (reference) position(s) the sequence position(s) in sequence coords are aligned to examples of ambiguous nucleotides in start/stop codons
ambgcd3c AMBIGUITY_IN_STOP_CODON sequence position(s) of stop codon model (reference) position(s) the sequence position(s) in sequence coords are aligned to examples of ambiguous nucleotides in start/stop codons
pepadjcy PEPTIDE_ADJACENCY_PROBLEM sequence position(s) of nucleotides inserted between two mature peptide predictions that are expected to be adjacent model (reference) position(s) corresponding to the end of the 5' mature peptide and the start of the 3' mature peptide (always length 2) mature peptide-specific alert examples
peptrans PEPTIDE_TRANSLATION_PROBLEM will be blank (-) will be blank (-) mature peptide-specific alert examples
lowsim5c, lowsim5n, lowsim5l LOW_FEATURE_SIMILARITY_START sequence position(s) at 5' end of sequence that have low similarity to the reference model (low similarity region overlaps with a predicted feature) model (reference) position(s) the sequence position(s) in sequence coords are aligned to examples of low similarity at start
lowsim5s LOW_SIMILARITY_START sequence position(s) at 5' end of sequence (not overlapping with a feature) that have low similarity to the reference model model (reference) position(s) the sequence position(s) in sequence coords are aligned to, or '-' if sequence is not aligned examples of low similarity at start
lowsim3c, lowsim3n, lowsim3l LOW_FEATURE_SIMILARITY_END sequence position(s) at 3' end of sequence that have low similarity to the reference model (low similarity region overlaps with a predicted feature) model (reference) position(s) the sequence position(s) in sequence coords are aligned to examples of low similarity at end
lowsim3s LOW_SIMILARITY_END sequence position(s) at 3' end of sequence (not overlapping with a feature) that have low similarity to the reference model model (reference) position(s) the sequence position(s) in sequence coords are aligned to, or '-' if sequence is not aligned examples of low similarity at end
lowsimic, lowsimin, lowsimil LOW_FEATURE_SIMILARITY sequence position(s) internal to a predicted feature (not including first or final position of the feature) that have low similarity to the reference model model (reference) position(s) the sequence position(s) in sequence coords are aligned to examples of internal low similarity
lowsimis LOW_SIMILARITY sequence position(s) internal to a sequence (not including first or final position of the sequence) and not overlapping with a feature that have low similarity to the reference model model (reference) position(s) the sequence position(s) in sequence coords are aligned to, or '-' if sequence is not aligned examples of internal low similarity
extrant5 EXTRA_SEQUENCE_START sequence position(s) at 5' end of sequence predicted to come before the first position of the reference model always 0..0:+ examples of extra sequence at start
extrant3 EXTRA_SEQUENCE_END sequence position(s) at 3' end of sequence predicted to come after the final position of the reference model always L..L:+, where L is the length of the reference model examples of extra sequence at end
deletins DELETION_OF_FEATURE will be blank (-) model (reference) positions that correspond to the feature that is deleted in the sequence deleted feature examples
deletinf DELETION_OF_FEATURE_SECTION will be blank (-) model (reference) positions that correspond to the segment of the feature that is deleted in the sequence deleted features examples
dupregin DUPLICATE_REGIONS N sets of sequence coordinates, in pairs, each pair is two hits that overlap in model coordinates, N will be a factor of 2 N model (reference) coordinates, one for each of the hits in the coverage determination stage that correspond to each set of sequence coordinates 1 to N duplicate regions example
discontn DISCONTINUOUS_SIMILARITY N sets of sequence coordinates, one for each hit in the coverage determination stage N model (reference) coordinates, one for each of the hits in the coverage determination stage that correspond to each set of sequence coordinates 1 to N discontinuous similarity example
indfstrn INDEFINITE_STRAND sequence coordinates of the best hit on the opposite strand from the overall best hit for this sequence in the coverage determination stage model (reference) coordinates for the hit pertaining to the sequence coordinates in sequence coords indefinite strand example
lowcovrg LOW_COVERAGE one or more set of sequence coordinates that are not covered by any hit to the model on the top-scoring strand in the coverage determination stage will be blank (-) low coverage example

toy50 toy model used in the examples of alert messages below

The toy50 model is a toy example used to illustrate many of the problems with sequences that VADR can detect using simple examples on this page. The toy50 model is 50 nucleotides long and includes 1 CDS feature from positions 11 to 31 with the name (product) of protein one . That CDS is composed of two adjacent mature peptides: protein one mp1 from positions 11 to 22 and protein one mp2 from positions 23 to 28. The final 3 nucleotides of the CDS, 29 to 31, are the stop codon. The model info file for the toy50 model is shown below.

MODEL toy50 cmfile:"toy50.cm" group:"toy" length:"50" subgroup:"A" blastdb:"toy50.protein.fa"
FEATURE toy50 type:"CDS" coords:"11..31:+" parent_idx_str:"GBNULL" gene:"one" product:"protein one"
FEATURE toy50 type:"mat_peptide" coords:"11..22:+" parent_idx_str:"0" product:"protein one mp1"
FEATURE toy50 type:"mat_peptide" coords:"23..28:+" parent_idx_str:"0" product:"protein one mp2"

The reference sequence for the toy50 model is shown below, as a Stockholm format alignment file (even though it has one sequence) with special markup in the form of #=GC columns to show where the CDS and mature peptide features are, as well as the sequence position information:

# STOCKHOLM 1.0

toy50              GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GC CDS1.11..31:+ ..........123123123123123123stp...................
#=GC MP1.11..22:+  ..........123123123123............................
#=GC MP2.23..28:+  ......................123123......................
#=GC COLX.         00000000011111111112222222222333333333344444444445
#=GC COL.X         12345678901234567890123456789012345678901234567890
//

Stockholm format is described in more detail at https://en.wikipedia.org/wiki/Stockholm_format and http://eddylab.org/infernal/Userguide.pdf (section 9: "File and output formats")


Examples of alerts and their corresponding .alt output lines

Frameshift example

relevant alert codes corresponding alert descriptions notes
fsthicft, fsthicfi POSSIBLE_FRAMESHIFT_HIGH_CONF only possible if --glsearch is not used
fstlocft, fstlocfi POSSIBLE_FRAMESHIFT_LOW_CONF only possible if --glsearch is not used
fstukcft, fstukcfi POSSIBLE_FRAMESHIFT only possible if --glsearch is used

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-frameshift.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-frameshift/va-example-frameshift.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq               ftr          ftr              ftr  alert           alert                               seq  seq       mdl  mdl  alert 
#idx   name       model  type         name             idx  code      fail  description                      coords  len    coords  len  detail
#----  ---------  -----  -----------  ---------------  ---  --------  ----  -----------------------------  --------  ---  --------  ---  ------
1.1.2  TOY50-FS1  toy50  CDS          protein_one        1  fsthicfi  yes   POSSIBLE_FRAMESHIFT_HIGH_CONF  13..25:+   13  14..23:+   10  high confidence possible frameshift in CDS (frame restored before end) [cause:insert,S:13..17(5),M:13; restore:delete,S:25,M:22..23(2); frame:1(3)1; length:3:(13):8; shifted_avgpp:0.825; exp_avgpp:0.892;]

Alignment of TOY50-FS1 sequence to the toy50 model: The output file va-example-frameshift/va-example-frameshift.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the frameshift, as explained more below. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

TOY50-FS1         -AAATCACCGATGcccccGTGATCGC--TACCATAAATGAGCATTCTACGTGCAT
#=GR TOY50-FS1 PP .**********987777789***998..59*************************
#=GC SS_cons      :::::::::::::.....:::::::::::::::::::::::::::::::::::::
#=GC RF           GAAATCACCGATG.....GTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.      0000000001111.....1111112222222222333333333344444444445
#=GC RFCOL.X      1234567890123.....4567890123456789012345678901234567890

How to interpret this alert based on the above output: As reported in the .alt file and .alt.list files as shown above, a possible frameshift exists in the CDS named protein one in the sequence named TOY50-FS1 which matches best to the model named toy50. The .alt file contains details on the frameshift. The frameshifted region is sequence positions 13 to 25 (seq coords: 13..25:+ in the .alt file) which is aligned to the reference model positions 14 to 23 (mdl coords: 14..23:+).

The alert detail field provides further information: the indels that "cause" the frameshifted region are an insertion of length 5 of nucleotides 13 to 17 after model position 13 (cause:insert,S:13..17(5),M:13;). The mutation that restores the frame is a deletion of length 2 after nucleotide 25 corresponding to model positions 22 and 23 (restore:delete:S:25,M:22..23(2);). The CDS starts in frame 1 for 3 nt, shifts into frame 3 for the frameshifted region for 13 nt and restores to frame 1 for 8 nt before the end of the CDS (frame:1(3)1 and length:3:(13):8). The table below show how to interpret different frame and length strings. The frame that the CDS starts in is defined as the 'expected frame' (this will always be 1 unless the CDS is truncated at the 5' end).

This frameshift is a high confidence frameshift in that the average posterior probability of the aligned nucleotides in the frameshifted region is 0.825 (shifted_frame:0.825) and of the expected region prior to (5' of) the frameshift is 0.892 (exp_frame:0.892) both of which exceed the threshold for high confidence (0.8 by default). (The PP of both the shifted and expected regions must exceed the high confidence threshold for a frameshift to be defined as high confidence.) Other possible frameshifts with lower posterior probability values will be reported with the POSSIBLE_FRAMESHIFT_LOW_CONF error. If the --glsearch option is used with v-annotate.pl, as is recommended with SARS-CoV-2 analysis, posterior probability values are not calculated and so all frameshifts are reported with the POSSIBLE_FRAMESHIFT error.

A separate alignment file showing the CDS features that include possible frameshifts can be optionally output from v-annotate.pl using the --out_fsstk option. An example excerpt from such an alignment file for this possible frameshift is below (see va-example-frameshift/va-exammle-frameshift.vadr.toy50.CDS.1.1.frameshift.stk). The #=GR PP shows an estimate of the posterior probability of each aligned nucleotide as explained more here. The #=GR CS line shows the implied frame of each aligned nucleotide and have i for inserted nucleotides and d for deleted reference positions. The #=GC RF line shows the reference model sequence.

TOY50-FS1         ATGCCCCCGTGATCGC--TACCATAA
#=GR TOY50-FS1 PP *987777789***998..59******
#=GR TOY50-FS1 CS 111iiiii33333333dd11111111
#=GC SS_cons      ::::::::::::::::::::::::::
#=GC RF           ATG.....GTGATCGCTTTACCATAA
#=GC RFCOLX.      111.....111111222222222233
#=GC RFCOL.X      123.....456789012345678901

Explanation of frame and length strings in POSSIBLE_FRAMESHIFT_* alert detail

frame string ......length_string...... 5' truncated? 3' truncated? frame of shifted region explanation
1(2) 10:(50) no no 2 frame 1 for 10 nt, frame 2 for 50 nt
<1(2) 10:(50) yes no 2 frame 1 for 10 nt, frame 2 for 50 nt
<1(2)> 10:(50) yes yes 2 frame 1 for 10 nt, frame 2 for 50 nt
1(23)131> 10:(50:30)20:9:31 no yes 2 then 3 frame 1 for 10 nt, frame 2 for 50 nt, frame 3 for 30 nt, frame 1 for 20 nt, frame 3 for 9 nt, frame 1 for 31 nt

Large insertion example

relevant alert codes corresponding alert descriptions
insertnn, insertnp INSERTION_OF_NT

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-insert.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-insert/va-example-insert.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq              ftr   ftr          ftr  alert           alert                 seq  seq       mdl  mdl  alert 
#idx   name      model  type  name         idx  code      fail  description        coords  len    coords  len  detail
#----  --------  -----  ----  -----------  ---  --------  ----  ---------------  --------  ---  --------  ---  ------
1.1.1  TOY50-I1  toy50  CDS   protein_one    1  insertnn  no    INSERTION_OF_NT  23..28:+    6  23..23:+    1  too large of an insertion in nucleotide-based alignment of CDS feature [6>2]

Alignment of TOY50-I1 sequence to the toy50 model: The output file va-example-insert/va-example-insert.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the insertion, as explained more below. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

TOY50-I1         -AAATCACCGATGGTGATCGCTTggggggTACCATAAATGAGCATTCTACGTGCAT
#=GR TOY50-I1 PP .********************9866666689*************************
#=GC SS_cons     :::::::::::::::::::::::......:::::::::::::::::::::::::::
#=GC RF          GAAATCACCGATGGTGATCGCTT......TACCATAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.     00000000011111111112222......222222333333333344444444445
#=GC RFCOL.X     12345678901234567890123......456789012345678901234567890
//

How to interpret this alert based on the above output: As reported in the .alt file shown above, in the CDS feature with name protein one, the nucleotides 23 to 28, length 6, (seq coords:23..28:+, seq len:6) on the + strand insert after reference model position 23 (mdl coords:23..23:+). This length exceeds the minimum allowed length of 2 (set with the v-annotate.pl option --nmaxins 2 option for purposes of this example, alert detail:[6>2]). You can see the gggggg insertion after model position 23 in the above alignment.

The insertnn alert is detected in the nucleotide alignment stage. A similar insertnp alert can be reported for insertions detected in the protein validation stage, and often you will see both alerts for the same insertion.


Large deletion example

relevant alert codes corresponding alert descriptions
deletinn, deletinp DELETION_OF_NT

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-delete.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-delete/va-example-delete.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq              ftr   ftr          ftr  alert           alert                seq  seq       mdl  mdl  alert 
#idx   name      model  type  name         idx  code      fail  description       coords  len    coords  len  detail
#----  --------  -----  ----  -----------  ---  --------  ----  --------------  --------  ---  --------  ---  ------
1.1.1  TOY50-D1  toy50  CDS   protein_one    1  deletinn  no    DELETION_OF_NT  15..15:+    1  17..19:+    3  too large of a deletion in nucleotide-based alignment of CDS feature [3>2]

Alignment of TOY50-D1 sequence to the toy50 model: The output file va-example-delete/va-example-delete.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the deletion, as explained more below. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

TOY50-D1         -AAATCACCGATGGTG---GCTTTACCATAAATGAGCATTCTACGTGCAT
#=GR TOY50-D1 PP .***********9987...89*****************************
#=GC SS_cons     ::::::::::::::::::::::::::::::::::::::::::::::::::
#=GC RF          GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.     00000000011111111112222222222333333333344444444445
#=GC RFCOL.X     12345678901234567890123456789012345678901234567890
//

How to interpret this alert based on the above output: As reported in the .alt file shown above, in the CDS feature with name protein one, the reference model positions 17 to 19, length 3 (mdl coords:17..19:+, mdl len:3) are deleted in the sequence TOY50-D1. The deletion occurs after position 15 (seq coords:15..15:+) in the sequence. The deletion length of 3 exceeds the minimum allowed length of 2 (set with the v-annotate.pl option --nmaxdel 2 option for purposes of this example, alert detail:[3>2]). You can see the three deleted positions (---) in model RF positions 17 to 19 in the above alignment.

The deletinn alert is detected in the nucleotide alignment stage. A similar deletnp alert can be reported for deletions detected in the protein validation stage, and often you will see both alerts for the same deletion.


Mutated start codon example

relevant alert code corresponding alert description
mutstart MUTATION_AT_START

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-start.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-start/va-example-start.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq              ftr          ftr          ftr  alert           alert                  seq  seq       mdl  mdl  alert 
#idx   name      model  type         name         idx  code      fail  description         coords  len    coords  len  detail
#----  --------  -----  -----------  -----------  ---  --------  ----  ----------------  --------  ---  --------  ---  ------
1.1.1  TOY50-S1  toy50  CDS          protein_one   1  mutstart  yes   MUTATION_AT_START  10..12:+    3  11..13:+    3  expected start codon could not be identified [ATT]

Alignment of TOY50-S1 sequence to the toy50 model: The output file va-example-start/va-example-start.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, allows you to see the alignment of the predicted start codon and surrounding sequence. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                           vvv
TOY50-S1         -AAATCACCGATTGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GR TOY50-S1 PP .*************************************************
#=GC SS_cons     ::::::::::::::::::::::::::::::::::::::::::::::::::
#=GC RF          GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.     00000000011111111112222222222333333333344444444445
#=GC RFCOL.X     12345678901234567890123456789012345678901234567890

How to interpret this alert based on the above output: The first three nucleotides of any CDS feature that is not truncated on the 5' end are checked to see if they are a valid start codon, and if not, the mutstart alert is reported. For this specific example, as reported in the .alt file shown above, the CDS start codon is from model (reference) positions 11 to 13 (mdl coords:11..13:+), and the first 3 nucleotides of the predicted CDS are positions 10 to 12 (seq coords:10..12:+). The predicted invalid ATT start codon (alert detail:[ATT]) can be seen in the above alignment, marked by the vvv characters at the top of the alignment. (These vvv characters have been added here, and do not exist in the actual output alignment file.)


Examples of stop codon problems

relevant alert codes corresponding alert descriptions
mutendcd, mutendns, mutendex MUTATION_AT_END
cdsstopn, cdsstopp CDS_HAS_STOP_CODON
unexleng UNEXPECTED_LENGTH

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-stop.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-stop/va-example-stop.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq               ftr   ftr          ftr  alert           alert                     seq  seq       mdl  mdl  alert 
#idx   name       model  type  name         idx  code      fail  description            coords  len    coords  len  detail
#----  ---------  -----  ----  -----------  ---  --------  ----  -------------------  --------  ---  --------  ---  ------
1.1.1  TOY50-SP1  toy50  CDS   protein_one    1  mutendcd  yes   MUTATION_AT_END      28..30:+    3  29..31:+    3  expected stop codon could not be identified, predicted CDS stop by homology is invalid [TAC]
1.1.2  TOY50-SP1  toy50  CDS   protein_one    1  mutendns  yes   MUTATION_AT_END             -    -         -    -  expected stop codon could not be identified, no in-frame stop codon exists 3' of predicted start codon [-]
#
2.1.1  TOY50-SP2  toy50  CDS   protein_one    1  mutendcd  yes   MUTATION_AT_END      27..29:+    3  28..30:+    3  expected stop codon could not be identified, predicted CDS stop by homology is invalid [ATA]
2.1.2  TOY50-SP2  toy50  CDS   protein_one    1  mutendex  yes   MUTATION_AT_END      31..33:+    3  33..35:+    3  expected stop codon could not be identified, first in-frame stop codon exists 3' of predicted stop position [TGA]
2.1.3  TOY50-SP2  toy50  CDS   protein_one    1  unexleng  yes   UNEXPECTED_LENGTH    10..29:+   20  11..30:+   20  length of complete coding (CDS or mat_peptide) feature is not a multiple of 3 [20]
#
3.1.1  TOY50-SP3  toy50  CDS   protein_one    1  mutendcd  yes   MUTATION_AT_END      28..30:+    3  29..31:+    3  expected stop codon could not be identified, predicted CDS stop by homology is invalid [TAC]
3.1.2  TOY50-SP3  toy50  CDS   protein_one    1  cdsstopn  yes   CDS_HAS_STOP_CODON   22..24:+    3  23..25:+    3  in-frame stop codon exists 5' of stop position predicted by homology to reference [TAA, shifted S:6,M:6]

Alignment of TOY50-SP1, TOY50-SP2 and TOY50-SP3 sequences to the toy50 model: The output file va-example-stop/va-example-stop.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the stop-codon related alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                                              vvv
TOY50-SP1         -AAATCACCGATGGTGATCGCTTTACCATACATGAGCATTCTACGTGCAT
#=GR TOY50-SP1 PP .*************************************************
TOY50-SP2         -AAATCACCGATGGTGATCGCTTTACCATA-CTGAGCATTCTACGTGCAT
#=GR TOY50-SP2 PP .***************************96.69*****************
TOY50-SP3         -AAATCACCGATGGTGATCGCTTAACCATACATGAGCATTCTACGTGCAT
#=GR TOY50-SP3 PP .*************************************************
#=GC SS_cons      ::::::::::::::::::::::::::::::::::::::::::::::::::
#=GC RF           GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.      00000000011111111112222222222333333333344444444445
#=GC RFCOL.X      12345678901234567890123456789012345678901234567890
                                        ^^^

How to interpret this alert based on the above output: Several checks are made on CDS features related to stop codons and these three sequences demonstrate failures of some of these checks.

  • TOY50-SP1 sequence:

The mutendcd alert with description MUTATION_AT_END is reported because it has an invalid stop codon TAC (alert detail:[TAC]) at positions 28 to 30 (seq coords:28..30:+) which are aligned to model positions 29 to 31 (mdl coords:29..31:+, marked by vvv characters in the alignment above), which are the final three positions of the predicted protein one CDS feature. The mutendns alert with description MUTATION_AT_END is also reported because no valid in-frame stop codon (in the same frame as the predicted start codon) exists in the remainder of the sequence.

  • TOY50-SP2 sequence:

The mutendcd alert with description MUTATION_AT_END is reported because it has an invalid stop codon ATA (alert detail:[ATA]) at positions 27 to 29 (seq coords:27..29:+) which are aligned to model positions 28 to 30 (mdl coords:28..30:+), which are the final 3 positions of the predicted protein one CDS feature. Unlike TOY50-SP1, there is an in-frame valid TGA stop codon 3' of the expected stop position (alert detail:[TGA]), at sequence positions 31 to 33 (seq coords:31..33:+) which align to model positions 33 to 35 (mdl coords:33..35:+), which cause a mutendex alert with description MUTATION_AT_END to be reported. Finally, because the predicted CDS feature is length 20 (alert detail:[20]), from sequence positions 10 to 29 (seq coords:10..29:+) aligned to model positions 11 to 30 (mdl coords:11..30:+), and 20 is not a multiple of 3, the unexleng alert with description UNEXPECTED_LENGTH is reported.

  • TOY50-SP3 sequence:

Like TOY50-SP1, the mutendcd alert with description MUTATION_AT_END is reported because it has an invalid stop codon TAC (alert detail:[TAC]) at positions 28 to 30 (seq coords:28..30:+) which are aligned to model positions 29 to 31 (mdl coords:29..31:+), which are the final three positions of the predicted protein one CDS feature. Unlike the other two sequences, there is an in-frame valid TAA stop codon 5' of the expected stop position (alert detail:[TAA), at sequence positions 22 to 24 (seq coords:22..24:+) which align to model positions 23 to 25 (mdl coords:23..25:+, marked by ^^^ characters at the bottom of the alignment above), which causes a cdsstopn alert with description CDS_HAS_STOP_CODON. The early stop codon shifts the stop codon 6 positions in the sequence and the reference relative to the expected stop position (alert detail:shifted S:6,M:6).

The cdsstopn alert is detected based on the nucleotide alignment. A similar cdsstopp alert with description CDS_HAS_STOP_CODON is reported when early in-frame stop codons are detected in the protein validation stage by blastx. Often you will see both alerts for the same early stop codon, but sometimes you will only see one or the other.


Examples of indefinite annotation at the start of a sequence or a feature

relevant alert codes corresponding alert descriptions
indf5gap, indf5lcc, indf5pst*, indf5lcn (not shown), indf5plg (not shown) INDEFINITE_ANNOTATION_START

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-indefstart.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-indefstart/va-example-indefstart.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq               ftr   ftr          ftr  alert           alert                             seq  seq       mdl  mdl  alert 
#idx   name       model  type  name         idx  code      fail  description                    coords  len    coords  len  detail
#----  ---------  -----  ----  -----------  ---  --------  ----  ---------------------------  --------  ---  --------  ---  ------
1.1.4  TOY50-IS1  toy50  CDS   protein_one    1  indf5gap  yes   INDEFINITE_ANNOTATION_START  10..10:+    1  11..11:+    1  alignment to homology model is a gap at 5' boundary [-]
#
2.1.1  TOY50-IS2  toy50  CDS   protein_one    1  indf5lcc  no    INDEFINITE_ANNOTATION_START    8..8:+    1  11..11:+    1  alignment to homology model has low confidence at 5' boundary for feature that is or matches a CDS [0.80<0.90]

Alignment of TOY50-IS1 and TOY50-IS2 sequences to the toy50 model: The output file va-example-indefstart/va-example-indefstart.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                            v
TOY50-IS1         -AAATCACCG-TGGTGATCGCTTTACCATAAATGAGCAT-----------
#=GR TOY50-IS1 PP .*******98.89**************************...........
TOY50-IS2         -AAATCAC--ATGGTGATCGCTTTACCATAAATGAGCAT-----------
#=GR TOY50-IS2 PP .*****98..89***************************...........
#=GC SS_cons      ::::::::::::::::::::::::::::::::::::::::::::::::::
#=GC RF           GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.      00000000011111111112222222222333333333344444444445
#=GC RFCOL.X      12345678901234567890123456789012345678901234567890

How to interpret these alerts based on the above output: Several checks are made on the first position of all features, and these two sequences demonstrate failures due to some of these checks.

  • TOY50-IS1 sequence:

The indf5gap alert with description INDEFINITE_ANNOTATION_START is reported because there is a gap at the start position (5' boundary) of the CDS named protein one. The first non-gap nucleotide in the predicted CDS is position 10 of the sequence (seq coords:10..10:+). The 5' boundary gap is position 11 in the model (mdl coords:11..11:+).

  • TOY50-IS2 sequence:

The indf5lcc alert with description INDEFINITE_ANNOTATION_START is reported because the posterior probability of the aligned nucleotide at the start position (5' boundary) of the CDS named protein one is too low. In this example the value is 0.8, and the minimum value to not report an alert is 0.9 (alert detail:[0.80<0.90]). The minimum is actually usually 0.8, but it was changed to 0.9 for demonstration purposes in this example with the option --indefann 0.9. The nucleotide aligned at the 5' boundary is sequence position 10 and it aligns to position 11 in the model. Posterior probabilities are explained more here.

A similar indf5lcn alert exists for non-coding (non-CDS and non-mature peptide) features, but no example is shown here. Having separate alerts for coding and non-coding features gives the user control over whether these types of alerts in coding versus non-coding features cause a sequence to fail or not.

  • Related indf5pst and indf5plg alerts (not shown in example sequences above):

If the protein validation stage for a sequence results in too short of blastx alignment of the predicted CDS to the reference protein that does not extend close enough to the 5' boundary of the predicted CDS, the indf5pst alert will be reported. This alert is not possible to generate with the toy50 toy model because the CDS in this example is too short for constructive use with blastx, but a fabricated example .alt file output line is shown below, in which the blastx alignment
extended only to sequence position 19, leaving positions 10 to 18 of the predicted CDS (seq coords:10..18:+), which align to model positions 11 to 19 (mdl coords:11..19:+), uncovered by the blastx alignment. This length difference of 9 exceeds the maximum allowed difference of 5 (alert detail:[9>5]), so the alert is reported. A similar indf5plg alert exists for when the blastx alignment extends longer than the predicted nucleotide alignment on the 5' end, but no example is shown here.

#      seq              ftr   ftr          ftr  alert           alert                             seq  seq       mdl  mdl  alert 
#idx   name      model  type  name         idx  code      fail  description                    coords  len    coords  len  detail
#----  --------- -----  ----  -----------  ---  --------  ----  ---------------------------  --------  ---  --------  ---  ------
4.2.1  TOY50.F1  toy50  CDS   protein_one    1  indf5pst  yes   INDEFINITE_ANNOTATION_START  10..18:+    9  11..19:+    9  protein-based alignment does not extend close enough to nucleotide-based alignment 5' endpoint [9>5]

Examples of indefinite annotation at end of a feature

relevant alert codes corresponding alert descriptions
indf3gap, indf3lcc, indf3pst*, indf3lcn (not shown), indf3plg (not shown) INDEFINITE_ANNOTATION_END

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-indefend.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-indefend/va-example-indefend.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq               ftr   ftr          ftr  alert           alert                             seq  seq       mdl  mdl  alert 
#idx   name       model  type  name         idx  code      fail  description                    coords  len    coords  len  detail
#----  ---------  -----  ----  -----------  ---  --------  ----  ---------------------------  --------  ---  --------  ---  ------
1.1.4  TOY50-IE1  toy50  CDS   protein_one    1  indf3gap  yes   INDEFINITE_ANNOTATION_END    28..28:+    1  31..31:+    1  alignment to homology model is a gap at 3' boundary [-]
#
2.1.1  TOY50-IE2  toy50  CDS   protein_one    1  indf3lcc  no    INDEFINITE_ANNOTATION_END    30..30:+    1  31..31:+    1  alignment to homology model has low confidence at 3' boundary for feature that is or matches a CDS [0.70<0.80]

Alignment of TOY50-IE1 and TOY50-IE2 sequences to the toy50 model: The output file va-example-indefend/va-example-indefend.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                                                v
TOY50-IE1         -AAATCACCGATGGTGATCGCTTTACCAT--ATGAGCAT-----------
#=GR TOY50-IE1 PP .**************************98..59******...........
TOY50-IE2         -AAATCACCGATGGTGATCCCTCTAGCATAGA-GAGCAT-----------
#=GR TOY50-IE2 PP .***************************9876.9*****...........
#=GC SS_cons      ::::::::::::::::::::::::::::::::::::::::::::::::::
#=GC RF           GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.      00000000011111111112222222222333333333344444444445
#=GC RFCOL.X      12345678901234567890123456789012345678901234567890

How to interpret these alerts based on the above output: Several checks are made on the first position of all features, and these two sequences demonstrate failures to some of these checks.

  • TOY50-IE1 sequence:

The indf3gap alert with description INDEFINITE_ANNOTATION_END is reported because there is a gap at the end position (3' boundary) of the CDS named protein one. The final non-gap nucleotide in the predicted CDS is position 28 (seq coords:28..28:+) of the sequence. The 3' boundary gap is position 31 in the model (mdl coords:31..31:+).

  • TOY50-IE2 sequence:

The indf3lcc alert with description INDEFINITE_ANNOTATION_END is reported because the posterior probability of the aligned nucleotide at the end position (3' boundary) of the CDS named protein one is too low. In this example the value is 0.7, and the minimum value to not report an alert is 0.8 (alert detail:[0.70<0.80]). The nucleotide aligned at the 3' boundary is sequence position 30 (seq coords:30..30:+) and it aligns to position 31 (mdl coords:31..31:+) in the model. Posterior probabilities are explained more here.

A similar indf3lcn alert exists for non-coding (non-CDS and non-mature peptide) features, but no example is shown here. Having separate alerts for coding and non-coding features gives the user control over whether these types of alerts in coding versus non-coding features cause a sequence to fail or not.

  • Related indf3pst and indf3plg alerts (not shown in example sequences above):

If the protein validation stage for a sequence results in too short of blastx alignment of the predicted CDS to the reference protein that does not extend close enough to the 3' boundary of the predicted CDS, the indf3pst alert will be reported. This alert is not possible to generate with the toy50 toy model because the CDS in this example is too short for constructive use with blastx. But a fabricated example .alt file output line is shown below, in which the blastx alignment
extended only to sequence position 21, leaving positions 22 to 30 of the predicted CDS (seq coords:22..30:+), which align to model positions 23 to 31 (mdl coords:23..31:+), uncovered by the blastx alignment. This length difference of 9 exceeds the maximum allowed difference of 8 (alert detail:[9>8]), so the alert is reported. A similar indf3plg alert exists for when the blastx alignment extends longer than the predicted nucleotide alignment on the 3' end, but no example is shown here.

#      seq              ftr   ftr          ftr  alert           alert                             seq  seq       mdl  mdl  alert 
#idx   name      model  type  name         idx  code      fail  description                    coords  len    coords  len  detail
#----  --------- -----  ----  -----------  ---  --------  ----  ---------------------------  --------  ---  --------  ---  ------
7.2.1  TOY50-F2  toy50  CDS   protein_one    1  indf3pst  yes   INDEFINITE_ANNOTATION_END    22..30:+    9  23..31:+    9  protein-based alignment does not extend close enough to nucleotide-based alignment 3' endpoint [9>8]

Examples of ambiguous nucleotides at the start of a sequence or a feature

relevant alert codes corresponding alert descriptions
ambgnt5s AMBIGUITY_AT_START
ambgnt5c AMBIGUITY_AT_CDS_START
ambgnt5f AMBIGUITY_AT_FEATURE_START

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-ambigstart.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-ambigstart/va-example-ambigstart.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq               ftr          ftr              ftr  alert           alert                            seq  seq       mdl  mdl  alert 
#idx   name       model  type         name             idx  code      fail  description                   coords  len    coords  len  detail
#----  ---------  -----  -----------  ---------------  ---  --------  ----  --------------------------  --------  ---  --------  ---  ------
1.1.1  TOY50-AS1  toy50  CDS          protein_one        1  ambgnt5c  no    AMBIGUITY_AT_CDS_START      10..13:+    4  11..14:+    4  first nucleotide of CDS is an ambiguous nucleotide [-]
1.2.1  TOY50-AS1  toy50  mat_peptide  protein_one_mp1    2  ambgnt5f  no    AMBIGUITY_AT_FEATURE_START  10..13:+    4  11..14:+    4  first nucleotide of non-CDS feature is an ambiguous nucleotide [-]
#
2.1.1  TOY50-AS2  toy50  -            -                  -  ambgnt5s  no    AMBIGUITY_AT_START           1..13:+   13   2..14:+   13  first nucleotide of the sequence is an ambiguous nucleotide [-]
2.2.1  TOY50-AS2  toy50  CDS          protein_one        1  ambgnt5c  no    AMBIGUITY_AT_CDS_START      10..13:+    4  11..14:+    4  first nucleotide of CDS is an ambiguous nucleotide [-]
2.3.1  TOY50-AS2  toy50  mat_peptide  protein_one_mp1    2  ambgnt5f  no    AMBIGUITY_AT_FEATURE_START  10..13:+    4  11..14:+    4  first nucleotide of non-CDS feature is an ambiguous nucleotide [-]

Alignment of TOY50-AS1 and TOY50-AS2 sequences to the toy50 model: The output file va-example-ambigstart/va-example-ambigstart.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                            vvvv
TOY50-AS1         -AAATCACCGNNNNTGATCGCTTTACCATAAATGAGCAT-----------
#=GR TOY50-AS1 PP .**************************************...........
TOY50-AS2         -NNNNNNNNNNNNNTGATCGCTTTACCATAAATGAGCAT-----------
#=GR TOY50-AS2 PP .**************************************...........
#=GC SS_cons      ::::::::::::::::::::::::::::::::::::::::::::::::::
#=GC RF           GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.      00000000011111111112222222222333333333344444444445
#=GC RFCOL.X      12345678901234567890123456789012345678901234567890

How to interpret these alerts based on the above output: If the first position of each sequence or predicted feature is an N, an alert is reported.

  • TOY50-AS1 sequence:

The ambgnt5c alert with description AMBIGUITY_AT_CDS_START is reported because the predicted CDS named protein one begins with 4 consecutive Ns from sequence positions 10 to 13 (seq coords:10..13:+) that align to the model (reference) positions 11 to 14 (mdl coords:11..14:+). These positions are marked with vvvv in the alignment above. This alert is specific to CDS features.

Similarly, the ambgnt5f alert with description AMBIGUITY_AT_FEATURE_START is reported because the predicted mat_peptide named protein one mp1 begins with 4 consecutive Ns from sequence positions 10 to 13 (seq coords:10..13:+) that align to the model (reference) positions 11 to 14 (mdl coords:11..14:+). These positions are marked with vvvv in the alignment above. This alert is only reported for non-CDS features.

  • TOY50-AS2 sequence:

This sequence is similar to TOY50-AS1 except the stretch of Ns begins at position 1 and extends to position 13 (seq coords:1..13:+). Because the first 4 nucleotides of the CDS and mat_peptide are Ns like in TOY50-AS1, the same ambgnt5c and ambgnt5f alerts are reported, but now with an additional ambgnt5s alert with description AMBIGUITY_AT_START because the beginning of the sequence begins with a consecutive string of 13 Ns.

This sequence is similar to TOY50-AS1 except the stretch of Ns begins at position 1 and extends to position 13 position of the sequence causing the ambgnt5s alert with description AMBIGUITY_AT_START to be reported because the sequence ends with a stretch of 13 consecutive Ns. Because the first 4 nucleotides of the CDS and mat_peptide are Ns like in TOY50-AS1, the same ambgnt5c and ambgnt5f alerts are reported as well.


Examples of ambiguous nucleotides at end of a sequence or a feature

relevant alert codes corresponding alert descriptions
ambgnt3s AMBIGUITY_AT_END
ambgnt3c AMBIGUITY_AT_CDS_END
ambgnt3f AMBIGUITY_AT_FEATURE_END

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-ambigend.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-ambigend/va-example-ambigend.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq               ftr   ftr          ftr  alert           alert                      seq  seq       mdl  mdl  alert 
#idx   name       model  type  name         idx  code      fail  description             coords  len    coords  len  detail
#----  ---------  -----  ----  -----------  ---  --------  ----  --------------------  --------  ---  --------  ---  ------
1.1.1  TOY50-AE1  toy50  CDS   protein_one    1  ambgnt3c  no    AMBIGUITY_AT_CDS_END  29..30:+    2  30..31:+    2  final nucleotide of CDS is an ambiguous nucleotide [-]
#
2.1.1  TOY50-AE2  toy50  -     -              -  ambgnt3s  no    AMBIGUITY_AT_END      29..38:+   10  30..39:+   10  final nucleotide of the sequence is an ambiguous nucleotide [-]
2.2.1  TOY50-AE2  toy50  CDS   protein_one    1  ambgnt3c  no    AMBIGUITY_AT_CDS_END  29..30:+    2  30..31:+    2  final nucleotide of CDS is an ambiguous nucleotide [-]

Alignment of TOY50-AE1 and TOY50-AE2 sequences to the toy50 model: The output file va-example-ambigend/va-example-ambigend.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                                               vv
TOY50-AE1         -AAATCACCGATGGTGATCGCTTTACCATNNATGAGCAT-----------
#=GR TOY50-AE1 PP .**************************************...........
TOY50-AE2         -AAATCACCGATGGTGATCGCTTTACCATNNNNNNNNNN-----------
#=GR TOY50-AE2 PP .**************************************...........
#=GC SS_cons      ::::::::::::::::::::::::::::::::::::::::::::::::::
#=GC RF           GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.      00000000011111111112222222222333333333344444444445
#=GC RFCOL.X      12345678901234567890123456789012345678901234567890

How to interpret these alerts based on the above output: If the first position of each sequence or predicted feature is an N, an alert is reported.

  • TOY50-AE1 sequence:

The ambgnt3c alert with description AMBIGUITY_AT_CDS_END is reported because the predicted CDS named protein one ends with 2 consecutive Ns from sequence positions 29 to 30 (seq coords:29..30:+) that align to the model (reference) positions 30 to 31 (mdl coords:30..31:+). These positions are marked with vv in the alignment above. This alert is specific to CDS features.

A similar ambgnt3f alert exists for non-coding (non-CDS and non-mature peptide) features, but no example is shown here. Having separate alerts for coding and non-coding features gives the user control over whether these types of alerts in coding versus non-coding features cause a sequence to fail or not.

  • TOY50-AE2 sequence:

This sequence is similar to TOY50-AE1 except the stretch of Ns begins at position 29 and continues to the position 38 (seq coords:29..38:+) which is the final position of the sequence causing the ambgnt3s alert with description AMBIGUITY_AT_END to be reported because the sequence ends with a stretch of 10 consecutive Ns. Because the final 2 nucleotides of the CDS are Ns like in TOY50-AS1, the same ambgnt3c alert is reported.


Examples of ambiguous nucleotides in start/stop codons that start/stop with canonical nt

relevant alert codes corresponding alert descriptions
ambgcd5c AMBIGUITY_IN_START_CODON
ambgcd3c AMBIGUITY_IN_STOP_CODON

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-ambigcodon.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-ambigend/va-example-ambigend.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq               ftr   ftr          ftr  alert           alert                      seq  seq       mdl  mdl  alert 
#idx   name       model  type  name         idx  code      fail  description             coords  len    coords  len  detail
#----  ---------  -----  ----  -----------  ---  --------  ----  --------------------  --------  ---  --------  ---  ------
1.1.1  TOY50-AC1  toy50  CDS   protein_one    1  ambgcd5c  no    AMBIGUITY_IN_START_CODON  10..12:+    3  11..13:+    3  5' complete CDS starts with canonical nt but includes ambiguous nt in its start codon [ANG]
#
2.1.1  TOY50-AC2  toy50  CDS   protein_one    1  ambgcd3c  no    AMBIGUITY_IN_STOP_CODON   28..30:+    3  29..31:+    3  3' complete CDS ends with canonical nt but includes ambiguous nt in its stop codon [TNA]

Alignment of TOY50-AC1 and TOY50-AC2 sequences to the toy50 model: The output file va-example-ambigcodon/va-example-ambigcodon.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                            vvv               
TOY50-AC1         -AAATCACCGANGNTGATCGCTTTACCATAAATGAGCAT-----------
#=GR TOY50-AC1 PP .**************************************...........
TOY50-AC2         -AAATCACCGATGGTGATCGCTTTACCATNAATGAGCAT-----------
#=GR TOY50-AC2 PP .**************************************...........
#=GC RF           GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.      00000000011111111112222222222333333333344444444445
#=GC RFCOL.X      12345678901234567890123456789012345678901234567890
                                              ^^^

How to interpret these alerts based on the above output: If the first position of each sequence or predicted feature is an N, an alert is reported.

  • TOY50-AC1 sequence:

The ambgcd5c alert with description AMBIGUITY_IN_START_CODON is reported because the predicted CDS named protein one begins with a start codon ANG that includes an ambiguous nucleotide but starts with a canonical nt at sequence positions 10 to 12 (seq coords:10..12:+) that align to the model (reference) positions 11 to 13 (mdl coords:11..13:+). These positions are marked with vv in the alignment above. This alert is specific to CDS features that are not truncated at the 5' end.

  • TOY50-AC2 sequence:

The ambgcd3c alert with description AMBIGUITY_IN_STOP_CODON is reported because the predicted CDS named protein one ends with a stop codon TNA that includes an ambiguous nucleotide but ends with a canonical nt at sequence positions 28 to 30 (seq coords:28..30:+) that align to the model (reference) positions 29 to 31 (mdl coords:29..31:+). These positions are marked with ^^ in the alignment above. This alert is specific to CDS features that are not truncated at the 3' end.


Examples of mature peptide-specific alerts

relevant alert codes corresponding alert descriptions
pepadjcy PEPTIDE_ADJACENCY_PROBLEM
peptrans PEPTIDE_TRANSLATION_PROBLEM

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-matpep.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-matpep/va-example-matpep.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq               ftr          ftr              ftr  alert           alert                             seq  seq       mdl  mdl  alert 
#idx   name       model  type         name             idx  code      fail  description                    coords  len    coords  len  detail
#----  ---------  -----  -----------  ---------------  ---  --------  ----  ---------------------------  --------  ---  --------  ---  ------
1.1.1  TOY50-MP1  toy50  mat_peptide  protein_one_mp1    2  pepadjcy  yes   PEPTIDE_ADJACENCY_PROBLEM    22..24:+    3  22..23:+    2  predictions of two mat_peptides expected to be adjacent are not adjacent [-]
#
2.1.1  TOY50-MP2  toy50  CDS          protein_one        1  mutstart  yes   MUTATION_AT_START            10..12:+    3  11..13:+    3  expected start codon could not be identified [GTG]
2.2.1  TOY50-MP2  toy50  mat_peptide  protein_one_mp1    2  peptrans  yes   PEPTIDE_TRANSLATION_PROBLEM         -    -         -    -  mat_peptide may not be translated because its parent CDS has a problem [-]
2.3.1  TOY50-MP2  toy50  mat_peptide  protein_one_mp2    3  peptrans  yes   PEPTIDE_TRANSLATION_PROBLEM         -    -         -    -  mat_peptide may not be translated because its parent CDS has a problem [-]

Alignment of TOY50-MP1 and TOY50-MP2 sequences to the toy50 model: The output file va-example-matpep/va-example-matpep.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                                       v   v
TOY50-MP1         -AAATCACCGATGGTGATCGCTgggTTACCATAAATGAGCAT-----------
#=GR TOY50-MP1 PP .******************99766589***************...........
TOY50-MP2         -AAATCACCGGTGGTGATCGCT...TTACCATAAATGAGCAT-----------
#=GR TOY50-MP2 PP .*********************...*****************...........
#=GC SS_cons      ::::::::::::::::::::::...::::::::::::::::::::::::::::
#=GC RF           GAAATCACCGATGGTGATCGCT...TTACCATAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.      0000000001111111111222...2222222333333333344444444445
#=GC RFCOL.X      1234567890123456789012...3456789012345678901234567890
                            ^^^

How to interpret these alerts based on the above output:

  • TOY50-MP1 sequence:

When the protein product of a CDS is cleaved into multiple mature peptides the mature peptide features typically are adjacent to each other like they are for the toy example toy50 (details here) in which protein one mp1 which spans nucleotide positions 11 to 22, is adjacent to protein one mp2 which spans positions 23 to 38. For these situations, v-annotate.pl checks that the predicted mature peptide features in each sequence are adjacent, and if not the pepadjcy alert is reported.

The pepadjcy alert with description PEPTIDE_ADJACENCY_PROBLEM is reported for the TOY50-MP1 sequence because the predicted CDS named protein one mp1 ends at position 21 and protein one mp2 begins at position 25, so the positions 22 to 24 lie between the two mature peptide predictions (seq coords:22..24:+). The model position that ends protein one mp1 is 22, and that begins protein one mp2 is 23 (mdl coords:22..23:+).

  • TOY50-MP2 sequence:

When a parent CDS of a mature peptide has a fatal alert, a peptrans alert with description PEPTIDE_TRANSLATION_PROBLEM is reported for all of the predicted child mature peptides. The TOY50-MP2 sequence shows an example, which has a mutstart alert for the CDS protein one which causes peptrans alerts for the two mature peptides.


Examples of low similarity to the model at the start of a sequence or a feature

relevant alert codes corresponding alert descriptions
lowsim5c, lowsim5n (not shown), lowsim5l (not shown) LOW_FEATURE_SIMILARITY_START
lowsim5s LOW_SIMILARITY_START

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-lowsimstart.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-lowsimstart/va-example-lowsimstart.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq                ftr   ftr          ftr  alert           alert                             seq  seq       mdl  mdl  alert 
#idx   name        model  type  name         idx  code      fail  description                    coords  len    coords  len  detail
#----  ----------  -----  ----  -----------  ---  --------  ----  ----------------------------  -------  ---  --------  ---  ------
1.2.3  TOY50-LSS1  toy50  CDS   protein_one    1  lowsim5c  no    LOW_FEATURE_SIMILARITY_START  1..13:+   13   5..13:+    9  region overlapping annotated feature that is or matches a CDS at 5' end of sequence lacks significant similarity [7 nt overlap b/t low similarity region of length 13 (1..13) and annotated feature (7..31)]
#
2.1.2  TOY50-LSS2  toy50  -     -              -  lowsim5s  yes   LOW_SIMILARITY_START          1..10:+   10   3..10:+    8  significant similarity not detected at 5' end of the sequence [low similarity region of length 10]

Alignment of TOY50-LSS1 and TOY50-LSS2 sequences to the toy50 model: The output file va-example-lowsimstart/va-example-lowsimstart.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                       vvvvvv  vvvvvvv                 
TOY50-LSS1         ----TTGTAG..GTTcgacGTGATCGCTTTACCATAAATGAGCAT-----------
#=GR TOY50-LSS1 PP ....998766..542223389************************...........
TOY50-LSS2         --GTTTAGTGgcATG....GTGATCGCTTTACCATAAATGAGCAT-----------
#=GR TOY50-LSS2 PP ..**9998775589*....**************************...........
#=GC SS_cons       ::::::::::..:::....:::::::::::::::::::::::::::::::::::::
#=GC RF            GAAATCACCG..ATG....GTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.       0000000001..111....1111112222222222333333333344444444445
#=GC RFCOL.X       1234567890..123....4567890123456789012345678901234567890

Explanation of lowsim5c alert: The 13 nucleotides from positions 1 to 13 in the sequence TOY50-LSS1 (seq coords:1..13:+) are at the 5' end of the sequence and overlap by 7 nucleotides with the predicted CDS protein one but are not similar to the reference model. These 13 nucleotides align to reference model positions 5 to 13 (mdl coords:5..13:+) and include an insertion of 4 nucleotides after reference position 13. The 13 nucleotides are marked by v characters in the alignment above.

Two similar alerts, lowsim5f and lowsim5l, exist for non-coding (non-CDS and non-mature peptide) features, but no examples are shown here. lowsim5f is for relatively short regions of low similarity, while lowsim5l is for longer regions (30nt or more). Having separate alerts for coding and non-coding features gives the user control over whether these types of alerts in coding versus non-coding features.

Explanation of lowsim5s alert: The first ten nucleotides in the sequence TOY50-LSS2 are not similar to the reference model (seq coords:1..10:+) and do not overlap with any predicted features. These ten nucleotides align to reference model positions 3 to 10, (mdl coords:3..10:+) and include two inserted nucleotides after reference position 10.

Regions of low similarity are detected in the coverage determination stage, as regions that are not covered by local alignment hits between the sequence and the model, not based on the global alignment determined in the alignment stage. In the alignment above, note that the nucleotides in the low similarity regions at the beginning of each sequence do not match well to the nucleotides in the reference model (#=GC RF line).


Examples of low similarity to the model at the end of a sequence or a feature

relevant alert codes corresponding alert descriptions
lowsim3c, lowsim3n (not shown), lowsim3l (not shown) LOW_FEATURE_SIMILARITY_END
lowsim3s LOW_SIMILARITY_END

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-lowsimend.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-lowsimend/va-example-lowsimend.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq                ftr          ftr              ftr  alert           alert                             seq  seq       mdl  mdl  alert 
#idx   name        model  type         name             idx  code      fail  description                    coords  len    coords  len  detail
#----  ----------  -----  -----------  ---------------  ---  --------  ----  ----------------------------  -------  ---  --------  ---  ------
1.2.2  TOY50-LSE1  toy50  CDS          protein_one        1  lowsim3c  no    LOW_FEATURE_SIMILARITY_END    26..37:+   12  26..31:+    6  region overlapping annotated feature that is or matches a CDS at 3' end of sequence lacks significant similarity [12 nt overlap b/t low similarity region of length 12 (26..37) and annotated feature (10..37)]
#
2.1.2  TOY50-LSE2  toy50  -            -                  -  lowsim3s  yes   LOW_SIMILARITY_END            40..53:+   14  35..46:+   12  significant similarity not detected at 3' end of the sequence [low similarity region of length 14]

Alignment of TOY50-LSE1 and TOY50-LSE2 sequences to the toy50 model: The output file va-example-lowsimend/va-example-lowsimend.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                                            vvvvvvvvvvvvv                 
TOY50-LSE1         -AAATCACCGATGGTGATCGCTTTACgtcccgtCTTAA----........---------------
#=GR TOY50-LSE1 PP .***********************975544444689**...........................
TOY50-LSE2         -AAATCACCGATGGTGATCGCTTTAC.......CATAAATGAcgatacacGAACTGCACGA----
#=GR TOY50-LSE2 PP .*************************.......********922222222334466799**....
#=GC SS_cons       ::::::::::::::::::::::::::.......:::::::::........:::::::::::::::
#=GC RF            GAAATCACCGATGGTGATCGCTTTAC.......CATAAATGA........GCATTCTACGTGCAT
#=GC RFCOLX.       00000000011111111112222222.......222333333........333344444444445
#=GC RFCOL.X       12345678901234567890123456.......789012345........678901234567890
                                                            ^^^^^^^^^^^^^^^^^^^^

Explanation of lowsim3c alert: The 12 nucleotides from positions 26 to 37 (seq coords:26..37:+) in the sequence TOY50-LSE1 are the 3' end of the sequence and overlap with the predicted CDS protein one but are not similar to the reference model. These 12 nucleotides align to reference model positions 26 to 31 (mdl coords:26..31:+) including an insertion of length 7 after model position 26. The 12 nucleotides are marked by v characters in the alignment above.

Two similar alerts, lowsim3f and lowsim3l, exist for non-coding (non-CDS and non-mature peptide) features, but no examples are shown here. lowsim3f is for relatively short regions of low similarity, while lowsim3l is for longer regions (30nt or more). Having separate alerts for coding and non-coding features gives the user control over whether these types of alerts in coding versus non-coding features.

Explanation of lowsim3s alert: The final 14 nucleotides from positions 40 to 53 (seq coords:40..53:+) in the sequence TOY50-LSE2 are not similar to the reference model and do not overlap with any predicted features. These 14 nucleotides align to reference model positions 35 to 46 (mdl coords:35..46:+), and include 8 inserted nucleotides after reference position 35. The alignment above shows the region of low similarity marked by ^ characters.

Regions of low similarity are detected in the coverage determination stage, as regions that are not covered by local alignment hits between the sequence and the model, not based on the global alignment determined in the alignment stage.


Examples of low similarity to the model in a region internal to a sequence or feature

relevant alert codes corresponding alert descriptions
lowsimic, lowsimin (not shown), lowsimil (not shown) LOW_FEATURE_SIMILARITY
lowsimis LOW_SIMILARITY

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-lowsimint.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-lowsimint/va-example-lowsimint.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq                ftr   ftr          ftr  alert           alert                         seq  seq       mdl  mdl  alert 
#idx   name        model  type  name         idx  code      fail  description                coords  len    coords  len  detail
#----  ----------  -----  ----  -----------  ---  --------  ----  ----------------------  ---------  ---  --------  ---  ------
1.2.3  TOY50-LSI1  toy50  CDS   protein_one  1    lowsimic  no    LOW_FEATURE_SIMILARITY  32..107:+   76  27..28:+    2  region overlapping annotated feature that is or matches a CDS lacks significant similarity [76 nt overlap b/t low similarity region of length 76 (32..107) and annotated feature (10..114)]

Alignment of TOY50-LSI1 sequence to the toy50 model: The output file va-example-lowsimint/va-example-lowsimint.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                                                  vvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv vvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv
TOY50-LSI1         -AAATCACCGATGGTGATCGCTTTACCaaagcagtacaggcacatgacaaagcagtacaggca-catgacaaagcagtacaggcacatgacaaagcagtacaggcacatgacaTAAATGAGCATTCTACGTGCAT
#=GR TOY50-LSI1 PP .**********************9875222222222222222222222222222222222222.222222222222222222222222222222222222222222222222267899*****************
#=GC SS_cons       :::::::::::::::::::::::::::....................................:.................................................::::::::::::::::::::::
#=GC RF            GAAATCACCGATGGTGATCGCTTTACC....................................A.................................................TAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.       000000000111111111122222222....................................2.................................................2333333333344444444445
#=GC RFCOL.X       123456789012345678901234567....................................8.................................................9012345678901234567890
//

Explanation of lowsimic alert: The 76 nucleotides from positions 32 to 107 (seq coords:32..107:+) in the sequence TOY50-LSI1 occur in the middle of the predicted CDS protein one but are not similar to the reference model. These 76 nucleotides align to reference model positions 27 to 28 (mdl coords:27..28:+), but are all actually inserted after those two reference positions as marked with v characters in the alignment shown above.

A similar lowsimin alert exists for non-coding (non-CDS and non-mature peptide) features, but no example is shown here. Having separate alerts for coding and non-coding features gives the user control over whether these types of alerts in coding versus non-coding

Two similar alerts, lowsimif and lowsimil, exist for non-coding (non-CDS and non-mature peptide) features, but no examples are shown here. lowsimif is for relatively short regions of low similarity, while lowsimil is for longer regions (30nt or more). Having separate alerts for coding and non-coding features gives the user control over whether these types of alerts in coding versus non-coding features.

Regions of low similarity are detected in the coverage determination stage, as regions that are not covered by local alignment hits between the sequence and the model, not based on the global alignment determined in the alignment stage. That is why the region marked with v characters above which corresponds to the region of low similarity does not include the full insertions after reference positions 27 and 28. Enough similarity exists in the beginning of the insertion after position 27 and at the end of the insertion after position 28 for a local alignment hit to include those regions.


Example of extra sequence at the start of a sequence

relevant alert codes corresponding alert descriptions
extrant5 EXTRA_SEQUENCE_START

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-extrant5.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-extrant5/va-example-extrant5.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq              ftr   ftr   ftr  alert           alert                    seq  seq     mdl  mdl  alert 
#idx   name      model  type  name  idx  code      fail  description           coords  len  coords  len  detail
#----  --------  -----  ----  ----  ---  --------  ----  --------------------  ------  ---  ------  ---  ------
1.1.1  TOY50-E5  toy50  -     -       -  extrant5  no    EXTRA_SEQUENCE_START  1..6:+    6  0..0:+    1  extra sequence detected 5' of expected sequence start [6 extra nucleotides]

Alignment of the TOY50-E5 sequence to the toy50 model: The output file va-example-extrant5/va-example-extrant5.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                 vvvvvv
TOY50-E5         ctacctGAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GR TOY50-E5 PP ********************************************************
#=GC SS_cons     ......::::::::::::::::::::::::::::::::::::::::::::::::::
#=GC RF          ......GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.     ......00000000011111111112222222222333333333344444444445
#=GC RFCOL.X     ......12345678901234567890123456789012345678901234567890

Explanation of extrant5 alert: The 6 nucleotides from positions 1 to 6 in the sequence TOY50-E5 (seq coords:1..6:+) are at the 5' end of the sequence and are aligned before the first model (nongap RF) position. The 6 nucleotides are marked by v characters in the alignment above.
By default if 5 or more nucleotides are inserted before the first model position a extrant5 alert is reported, but 5 is changeable to <n> with the --extrant5 <n> option to v-annotate.pl.


Example of extra sequence at the start of a sequence

relevant alert codes corresponding alert descriptions
extrant3 EXTRA_SEQUENCE_END

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-extrant3.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-extrant3/va-example-extrant3.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq              ftr   ftr   ftr  alert           alert                    seq  seq       mdl  mdl  alert 
#idx   name      model  type  name  idx  code      fail  description           coords  len    coords  len  detail
#----  --------  -----  ----  ----  ---  --------  ----  ------------------  --------  ---  --------  ---  ------
1.1.1  TOY50-E3  toy50  -     -       -  extrant3  no    EXTRA_SEQUENCE_END  51..59:+    9  50..50:+    1  extra sequence detected 3' of expected sequence end [9 extra nucleotides]

Alignment of the TOY50-E3 sequence to the toy50 model: The output file va-example-extrant3/va-example-extrant3.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                                                                   vvvvvvvvv
TOY50-E3         GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCATcaaaaaaaa
#=GR TOY50-E3 PP *************************************************9*********
#=GC SS_cons     ::::::::::::::::::::::::::::::::::::::::::::::::::.........
#=GC RF          GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT.........
#=GC RFCOLX.     00000000011111111112222222222333333333344444444445.........
#=GC RFCOL.X     12345678901234567890123456789012345678901234567890.........

Explanation of extrant3 alert: The 9 nucleotides from positions 51 to 59 in the sequence TOY50-E3 (seq coords:51..59:+) are at the 3' end of the sequence and are aligned after the final model (nongap RF) position. The 9 nucleotides are marked by v characters in the alignment above.
By default if 5 or more nucleotides are inserted after the final model position a extrant3 alert is reported, but 5 is changeable to <n> with the --extrant3 <n> option to v-annotate.pl.


Example of a deleted feature

relevant alert codes corresponding alert descriptions
deletins, deletina (not shown) DELETION_OF_FEATURE
deletinf (not shown) DELETION_OF_FEATURE_SECTION

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-delftr.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-delftr/va-example-delftr.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq               ftr   ftr   ftr  alert           alert                   seq  seq       mdl  mdl  alert 
#idx   name       model  type  name  idx  code      fail  description          coords  len    coords  len  detail
#----  ---------  -----  ----  ----  ---  --------  ----  -------------------  ------  ---  --------  ---  ------
1.1.1  TOY50-DF1  toy50  -     -       -  deletins  yes   DELETION_OF_FEATURE       -    -  23..28:+    6  internal deletion of a complete feature [mat_peptide feature number 2: protein one mp2]

Alignment of TOY50-DF1 sequence to the toy50 model: The output file va-example-delftr/va-example-delftr.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                                        vvvvvv
TOY50-DF1         -AAATCACCGATGGTGATCGC--------AAATGAGCATTCTACGTGCAT
#=GR TOY50-DF1 PP .*****************997........79*******************
#=GC SS_cons      ::::::::::::::::::::::::::::::::::::::::::::::::::
#=GC RF           GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.      00000000011111111112222222222333333333344444444445
#=GC RFCOL.X      12345678901234567890123456789012345678901234567890

Explanation of deletins alert: The mature peptide feature named protein one mp2 is deleted in the sequence TOY50-DF1, as indicated in the alert detail field of the .alt file above. That mature peptide spans model positions 23 to 28 (model coords:23..28:+) which can be seen as gaps in the above alignment, and are marked with positions v.

A similar deletina alert will be reported if the feature that is deleted has the is_deletable field set to 1 in the input .minfo file. No example of this alert is shown here.

A similar deletinf alert exists for multi-segment features for which one or more but not all segments are deleted. The associated value for this alert in the alert desc field is DELETION_OF_FEATURE_SECTION. No example of this alert is shown here.


Example of duplicate regions

relevant alert codes corresponding alert descriptions
dupregin DUPLICATE_REGION

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-dupregin.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-dupregin/va-example-dupregin.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq               ftr   ftr   ftr  alert           alert                           seq  seq              mdl  mdl  alert 
#idx   name       model  type  name  idx  code      fail  description                  coords  len           coords  len  detail
#----  ---------  -----  ----  ----  ---  --------  ----  -----------------  ----------------  ---  ---------------  ---  ------
1.1.1  TOY50-DR1  toy50  -     -       -  dupregin  yes   DUPLICATE_REGIONS  1..49:+,50..86:+   86  2..50:+,7..43:+   86  similarity to a model region occurs more than once [7..43 (len 37>=20) hits 1 (39.5 bits) and 2 (27.0 bits)]

Alignment of TOY50-DR1 sequence to the toy50 model: The output file va-example-dupregin/va-example-dupregin.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                                                                    |positions 50 to 86 in the sequence-|
                                                                    vvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv
TOY50-DR1         -AAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCataccgatggtgatcgctttaccataaatgagcattctA-
#=GR TOY50-DR1 PP .********************************************987444444444444444444444444444444444444446.
#=GC SS_cons      ::::::::::::::::::::::::::::::::::::::::::::::::......................................::
#=GC RF           GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGC......................................AT
#=GC RFCOLX.      000000000111111111122222222223333333333444444444......................................45
#=GC RFCOL.X      123456789012345678901234567890123456789012345678......................................90
                        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                        |--positions 7 to 43 in the model---|

Explanation of dupregin alert: There are two hits in the coverage determination stage, which overlap in model coordinates by more than 20 positions which cause the dupregin alert to be reported. Specifically, hit 1 is from positions 1 to 49 in the sequence (seq coords:1..49:+) and 2 to 50 in the model (mdl coords:2..50:+) and hit 2 is from positions 50 to 86 in the sequence (seq coords:50..86:+) and positions 7 to 43 in the model (mdl coords;7..43)). The hits overlap by 37 model positions from 7 to 43, which is above the minimum length threshold for reporting this alert of 20 (alert detail: 7..43 (len 37 >= 20)). The first hit has a bit score of 39.5 bits and the second hit of 27.0 bits (alert detail: hits 1 (39.5 bits) and 2 (27.0 bits)).

In the alignment of the sequence above sequence positions 50 to 86 are marked marked by v characters at the top of the alignment, which match identically to positions 7 to 43 in the reference model, marked by ^ at the bottom of the alignment.


Example of discontinuous similarity

relevant alert codes corresponding alert descriptions
discontn *DISCONTINUOUS_SIMILARITY

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-discontn.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-discontn/va-example-discontn.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq               ftr   ftr   ftr  alert           alert                                  seq  seq               mdl  mdl  alert 
#idx   name       model  type  name  idx  code      fail  description                         coords  len            coords  len  detail
#----  ---------  -----  ----  ----  ---  --------  ----  ------------------------  ----------------  ---  ----------------  ---  ------
1.1.1  TOY50-DS1  toy50  -     -       -  discontn  yes   DISCONTINUOUS_SIMILARITY  1..26:+,27..49:+   49  25..50:+,2..24:+   49  not all hits are in the same order in the sequence and the homology model [seq order: 1,2, mdl order: 2,1]

Alignment of TOY50-DS1 sequence to the toy50 model: The output file va-example-discontn/va-example-discontn.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                                                                    |-27 to 49 in the seq-|
                                                                    vvvvvvvvvvvvvvvvvvvvvvv
TOY50-DS1         ------------------------ACCATAAATGAGCATTCTACGTGCAtaaatcaccgatggtgatcgcttT
#=GR TOY50-DS1 PP ........................**********************98666666667777777777777777*
#=GC SS_cons      :::::::::::::::::::::::::::::::::::::::::::::::::.......................:
#=GC RF           GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCA.......................T
#=GC RFCOLX.      0000000001111111111222222222233333333334444444444.......................5
#=GC RFCOL.X      1234567890123456789012345678901234567890123456789.......................0
                   ^^^^^^^^^^^^^^^^^^^^^^^
                   |--2-24 in the model--|

Explanation of discontn alert: For the TOY50-DS1 sequence, there are two hits in the coverage determination stage and they are not in the same order in sequence and model coordinates. The positional information for these hits are reported in the seq coords, mdl coords and alert detail fields of the .alt file above. Specifically, hit 1 is from positions 1 to 26 in the sequence (seq coords:1..26:+) and 25 to 50 in the model (mdl coords:25..50:+) and hit 2 is from positions 27 to 49 in the sequence (seq coords:27..49:+) and positions 2 to 24 in the model (mdl coords:2..24:+). So hit 1 comes before hit 2 in the sequence, but hit 2 comes before hit 1 in the model. The order of the hits in the sequence and model is reported in the alert detail field: seq order: 1,2, mdl order: 2,1.

In the alignment of the sequence above, sequence positions 27 to 49 are marked marked by v characters at the top of the alignment. These sequence positions match identically to positions 2 to 24 in the reference model, marked by ^ at the bottom of the alignment, showing why the second hit is to the end of the sequence and the beginning of the model. The first hit pertains to the the beginning of the sequence which matches well to the end of the model (the region with * values in the PP line).


Example of indefinite strand

relevant alert codes corresponding alert descriptions
indfstrn, indfstrp (not shown) INDEFINITE_STRAND

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-indfstrn.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-indfstrn/va-example-indfstrn.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq               ftr   ftr   ftr  alert           alert                   seq  seq       mdl  mdl  alert 
#idx   name       model  type  name  idx  code      fail  description          coords  len    coords  len  detail
#----  ---------  -----  ----  ----  ---  --------  ----  -----------------  --------  ---  --------  ---  ------
1.1.2  TOY50-IS1  toy50  -     -       -  indfstrn  yes   INDEFINITE_STRAND  49..25:-   25  26..50:+   25  significant similarity detected on both strands [score:12.7>12.0]

Alignment of TOY50-IS1 sequence to the toy50 model: The output file va-example-indfstrn/va-example-indfstrn.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                                           |  25 to 49 in the seq  |
                                           vvvvvvvvvvvvvvvvvvvvvvvvv
TOY50-IS1         -AAATCACCGATGGTGATCGCTTTAATGCAcgtagAATGCTCATTTATGG-----
#=GR TOY50-IS1 PP .**********************98544332222389999**********.....
#=GC SS_cons      ::::::::::::::::::::::::::::::.....::::::::::::::::::::
#=GC RF           GAAATCACCGATGGTGATCGCTTTACCATA.....AATGAGCATTCTACGTGCAT
#=GC RFCOLX.      000000000111111111122222222223.....33333333344444444445
#=GC RFCOL.X      123456789012345678901234567890.....12345678901234567890
                                           ^^^^^     ^^^^^^^^^^^^^^^^^^^^
                                           |   26 to 50 in the model    |

Explanation of indfstrn alert: For the TOY50-IS1 sequence there are two hits in the coverage determination stage and they are not to the same strand. The positional information for these hits are reported in the seq coords, mdl coords and alert detail fields of the .alt file above. Specifically, the top-scoring hit is to the + strand, but the second hit with a score of 12.7 bits is on the - strand. The second hit occurs from positions to 49 to 25 on the negative strand (seq coords:49..25:-) to positions 26 to 50 in the reference model (mdl coords:26..50:+) The second hit therefore conflicts with the top-scoring hit and so it is the one that is described in the seq coords and mdl coords fields. Normally, a indfstrn alert is only reported if a hit on the opposite strand from the top hit exceeds 25 bits, but to construct this example, the minimum of 25 bits was lowered to 12 bits with the --indefstr 12 flag to v-annotate.pl.

In the alignment of the sequence above, sequence positions 25 to 49 are marked marked by v characters at the top of the alignment, after reverse complementing these match identically to positions 26 to 50 in the reference model, marked by ^ at the bottom of the alignment.

The indftrsn alert is detected in the coverage determination stage which is based on the nucleotide sequence. A similar indfstrp alert can be reported when blastx hits occur on both strands as detected in the protein validation stage, but is not shonw. Sometimes you will see both alerts for the same sequence.


Example of low coverage

relevant alert codes corresponding alert descriptions
lowcovrg LOW_COVERAGE

Instructions to reproduce this example and create the files discussed below:

> sh $VADRSCRIPTSDIR/documentation/alert-files/example-lowcovrg.sh

The above command will run v-annotate.pl and create many output files in a newly created directory. The output shown in the box below is from the .alt output file (va-example-lowcovrg/va-example-lowcovrg.vadr.alt) . Similar information with the seq and mdl coords fields can be found in the .alt.list output file and in the detailed error report .tsv file generated by the GenBank submission portal. You may have to scroll to the right to see the alert detail field.

Example lines from .alt file:

#      seq               ftr   ftr   ftr  alert           alert                                  seq  seq               mdl  mdl  alert 
#idx   name       model  type  name  idx  code      fail  description                         coords  len            coords  len  detail
#----  ---------  -----  ----  ----  ---  --------  ----  ------------------------  ----------------  ---  ----------------  ---  ------
1.1.1  TOY50-LC1  toy50  -     -       -  lowcovrg  yes   LOW_COVERAGE          1..25:+   25       -    -  low sequence fraction with significant similarity to homology model [0.545<0.900]

Alignment of TOY50-LC1 sequence to the toy50 model: The output file va-example-lowcovrg/va-example-lowcovrg.vadr.toy50.align.stk. includes the alignment shown below. Looking at this alignment, or an alignment of the sequence generated by a different program, can be helpful in understanding the alerts. The #=GC RF line shows the toy50 reference model sequence. The #=GR PP line indicates confidence estimates for each aligned nucleotide as explained more here. This alignment is only output when the --keep or --out_stk options are used with v-annotate.pl.

                  |1 to 25 in the sequence|
                  vvvvvvvvvvvvvvvvvvvvvvvvv
TOY50-LC1         GAtacatcgatcatcaatccgggacaAATCACCGATGGTGATCGCTTTACCATAA-------------------
#=GR TOY50-LC1 PP *64444444444444333333333336899*************************...................
#=GC SS_cons      ::........................::::::::::::::::::::::::::::::::::::::::::::::::
#=GC RF           GA........................AATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GC RFCOLX.      00........................000000011111111112222222222333333333344444444445
#=GC RFCOL.X      12........................345678901234567890123456789012345678901234567890

Explanation of lowcovrg alert: For the TOY50-LC1 sequence the nucleotides from 1 to 25 are not part of any hit to the model in the coverage determation stage (seq coords:1..25:+). These 25 nucleotides make up 45.5% the total length of the sequence, meaning that only 54.5% of the sequence is covered by hits to the model (alert detail:[0.545<0.900]). The lowcovrg alert is reported for any sequence if less than 90% of the sequence is covered by hits to the model (changeable to <x> with the --lowcov <x> option).

In the alignment of the sequence above, note that positions 1 to 25, marked by v characters at the top of the alignment, do not match well to the model.


Toy example of a model: TOY50

The TOY50 model is a toy example used to illustrate many of the problems with sequences that VADR can detect using simple examples on this page. The TOY50 model is 50 nucleotides long and includes 1 CDS feature from positions 11 to 31 with the name (product) of protein one . That CDS is composed of two adjacent mature peptides: protein one mp1 from positions 11 to 22 and protein one mp2 from positions 23 to 28. The final 3 nucleotides of the CDS, 29 to 31, are the stop codon. The model info file for the TOY50 model is shown below.

MODEL TOY50 cmfile:"toy50.cm" group:"toy" length:"50" subgroup:"A" blastdb:"toy50.protein.fa"
FEATURE TOY50 type:"CDS" coords:"11..31:+" parent_idx_str:"GBNULL" gene:"one" product:"protein one"
FEATURE TOY50 type:"mat_peptide" coords:"11..22:+" parent_idx_str:"0" product:"protein one mp1"
FEATURE TOY50 type:"mat_peptide" coords:"23..28:+" parent_idx_str:"0" product:"protein one mp2"

The reference sequence for the TOY50 model is shown below, as a Stockholm format alignment file (even though it has one sequence) with special markup in the form of #=GC columns to show where the CDS and mature peptide features are, as well as the sequence position information:

# STOCKHOLM 1.0

TOY50              GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCTACGTGCAT
#=GC CDS1.11..31:+ ..........123123123123123123stp...................
#=GC MP1.11..22:+  ..........123123123123............................
#=GC MP2.23..28:+  ......................123123......................
#=GC RFCOLX.       00000000011111111112222222222333333333344444444445
#=GC RFCOL.X       12345678901234567890123456789012345678901234567890
//

Stockholm format is described in more detail at https://en.wikipedia.org/wiki/Stockholm_format and http://eddylab.org/infernal/Userguide.pdf (section 9: "File and output formats")


Posterior probability annotation in VADR output Stockholm alignments

The v-annotate.pl script uses a probabilistic model called a covariance model (CM) to calculate glocal alignments of input sequences in the alignment stage (unless the --glsearch option is used, in which case a non-probabilistic alignment algorithm is used). As part of the CM alignment calculation, the posterior probability (PP) that each aligned nucleotide appears at its assigned alignment position is calculated. These posterior probabilities are confidence estimates that each nucleotide is correctly aligned given the parameters of the CM. The Stockholm alignments output from v-annotate.pl include annotation on the alignment that indicates these PP values. An example is below:

# STOCKHOLM 1.0
#=GF AU Infernal 1.1.4

toy50-1         GGTATCACAGATGGGATCCGCTGACTCAT-AATGTGTGTTCAaAAGTGCAT
#=GR toy50-1 PP 999*********99888888877766766.567888887775257999999
#=GC RF         GAAATCACCGATGGTGATCGCTTTACCATAAATGAGCATTCT.ACGTGCAT
#=GC RFCOLX.    000000000111111111122222222223333333333444.44444445
#=GC RFCOL.X    123456789012345678901234567890123456789012.34567890

The top line beginning with toy50-1 shows the alignment of the sequence named toy50-1. - characters here indicate deletions with respect to the model (gaps).

The #=GC RF line shows the reference model sequence. . characters in this line indicate positions that are insertions relative to the model in one or more of the aligned sequences.

The #=GC RFCOLX. line shows the tens value of the reference position of each column. The #=GC RFCOL.X line shows the ones value of the reference position of each column. (E.g. the final column is reference position 50.)

The posterior probability confidence estimates are shown in the #=GR toy50-1 PP line. Characters in PP rows have 12 possible values:

PP character meaning
. this position is a gap in the sequence
0 posterior probability between 0.00 and 0.05
1 posterior probability between 0.05 and 0.15
2 posterior probability between 0.15 and 0.25
3 posterior probability between 0.25 and 0.35
4 posterior probability between 0.35 and 0.45
5 posterior probability between 0.45 and 0.55
6 posterior probability between 0.55 and 0.65
7 posterior probability between 0.65 and 0.75
8 posterior probability between 0.75 and 0.85
9 posterior probability between 0.85 and 0.95
* posterior probability between 0.95 and 1.00

In the above alignment, the 9 positions 4 to 12 have * PP values, indicating that those positions are very confidently aligned at the correct positions given the parameters of the model. As you might expect, these positions nearly exactly match the model nucleotides they are aligned to.

Positions surrounding insertions and deletions tend to have lower PP values because there is more uncertainty as to where each specific nucleotide should be placed in the alignment. For example, reference position 29 and 30 which are separated by an insertion in the sequence have PP values of 6 and 5.

PP values are used by v-annotate.pl in two contexts:

  1. To distinguish between low-confidence and high-confidence frameshift regions (fsthicf* vs fstlocf* alerts, see this example).
  2. To report alerts when feature boundaries have low confidence (indf5lc* and indf3lc* alerts, see example at 5' end and 3' end.

Questions, comments or feature requests? Send a mail to eric.nawrocki@nih.gov.