The format is based on Keep a Changelog
and this project adheres to Semantic Versioning.
- #625 - Updated link to geNomad database for downloading (reported by @amizeranschi, fix by @jfy133)
- #618 - Fix CENTRIFUGE mkfifo failures by using work directory /tmp (reported by @skrakau, fix by @jfy133)
| Tool |
Previous version |
New version |
| Centrifuge |
1.0.4_beta |
1.0.4.1 |
- #615 - Add new logo (by @jfy133)
- #599 - Update to nf-core v2.13.1
TEMPLATE (by @jfy133)
- #614 - Update to nf-core v2.14.1
TEMPLATE (by @jfy133)
- #606 - Prevent pipeline crash when premade mashdb given to or no alignments found with GTDB-TK_CLASSIFYWF (reported by @cedwardson4, fix by @jfy133)
- #599 - Direct reads input (
--input 'sample_{R1,R2}.fastq.gz') is no longer supported, all input must come via samplesheets (by @jfy133)
- #581 - Added explicit licence text to headers of all custom scripts (reported by @FriederikeHanssen and @maxibor, fix by @jfy133)
- #602 - Co-binning when using aDNA mode now enabled (added by @maxibor)
- #583 - Fix GTDB database input when directory supplied (fix by @jfy133)
- #575 - Deactivated MetaSPAdes, Centrifuge, and GTDB in test_full profile due to some container incompatibilities in nf-core megatest AWS configurations (by @jfy133)
- #574 - Fix wrong channel going to BIN_SUMMARY (fix by @maxibor)
- #562 - Add CAT summary into the global bin_summary (by @maxibor)
- #565 - Add warning of empty GTDB-TK results if no contigs pass completeness filter (by @jfy133 and @maxibor)
- #563 - Update to nf-core v2.12
TEMPLATE (by @CarsonJM)
- #566 - More logical ordering of MultiQC sections (assembly and bin sections go together respectively) (fix by @jfy133)
- #548 - Fixes to (reported by @maxibor, @PPpissar, @muniheart, @llborcard, fix by @maxibor)
- GTDBK-TK execution
- CAT/QUAST/DEPTH bin summary file name collisions
- BUSCO database parsing
- Correct CAT name files
- #558 - Fix bug in run merging when dealing with single end data (reported by @roberta-davidson, fix by @jfy133)
- #489 - Fix file name collision clashes for CHECKM, CAT, GTDBTK, and QUAST (reported by @tillenglert and @maxibor, fix by @maxibor)
- #533 - Fix glob pattern for publishing MetaBAT2 bins in results (reported by @patriciatran, fix by @jfy133)
- #535 - Fix input validation pattern to again allow direct FASTQ input (reported by @lennijusten, @emnilsson, fix by @jfy133, @d4straub, @mahesh-panchal, @nvnieuwk)
| Tool |
Previous version |
New version |
| CAT |
4.6 |
5.2.3 |
- #536 - Remove custom function with native Nextflow for checking file extension (reported by @d4straub, fix by @jfy133)
- #504 - New parameters
--busco_db, --kraken2_db, and --centrifuge_db now support directory input of a pre-uncompressed database archive directory (by @gregorysprenger).
- #511 - Update to nf-core 2.10
TEMPLATE (by @jfy133)
- #504 -
--save_busco_reference is now replaced by --save_busco_db (by @gregorysprenger).
- #514 - Fix missing CONCOCT files in downstream output (reported by @maxibor, fix by @jfy133)
- #515 - Fix overwriting of GUNC output directories when running with domain classification (reported by @maxibor, fix by @jfy133)
- #516 - Fix edge-case bug where MEGAHIT re-uses previous work directory on resume and fails (reported by @husensofteng, fix by @prototaxites)
- #520 - Fix missing Tiara output files (fix by @jfy133)
- #522 - Fix 'nulls' in depth plot PNG files (fix by @jfy133)
- #504 -
--busco_reference, --busco_download_path, --save_busco_reference parameters have been deprecated and replaced with new parameters (by @gregorysprenger).
- #497 - Adds support for pointing at a local db for krona, using the parameter
--krona_db (by @willros).
- #395 - Adds support for fast domain-level classification of bins using Tiara, to allow bins to be separated into eukaryotic and prokaryotic-specific processes.
- #422 - Adds support for normalization of read depth with BBNorm (added by @erikrikarddaniel and @fabianegli)
- #439 - Adds ability to enter the pipeline at the binning stage by providing a CSV of pre-computed assemblies (by @prototaxites)
- #459 - Adds ability to skip damage correction step in the ancient DNA workflow and just run pyDamage (by @jfy133)
- #364 - Adds geNomad nf-core modules for identifying viruses in assemblies (by @PhilPalmer and @CarsonJM)
- #481 - Adds MetaEuk for annotation of eukaryotic MAGs, and MMSeqs2 to enable downloading databases for MetaEuk (by @prototaxites)
- #437 -
--gtdb_db also now supports directory input of an pre-uncompressed GTDB archive directory (reported by @alneberg, fix by @jfy133)
- #494 - Adds support for saving the BAM files from Bowtie2 mapping of input reads back to assembly (fix by @jfy133)
- #428 #467 - Update to nf-core 2.8, 2.9
TEMPLATE (by @jfy133)
- #429 - Replaced hardcoded CheckM database auto-download URL to a parameter (reported by @erikrikarddaniel, fix by @jfy133)
- #441 - Deactivated CONCOCT in AWS 'full test' due to very long runtime (fix by @jfy133).
- #442 - Remove warning when BUSCO finds no genes in bins, as this can be expected in some datasets (reported by @Lumimar, fix by @jfy133).
- #444 - Moved BUSCO bash code to script (by @jfy133)
- #477 -
--gtdb parameter is split into --skip_gtdbtk and --gtdb_db to allow finer control over GTDB database retrieval (fix by @jfy133)
- #500 - Temporarily disabled downstream processing of both refined and raw bins due to bug (by @jfy133)
- #496 - Fix help text for paramters
--bowtie2_mode, spades_options and megahit_options (by @willros)
- #400 - Fix duplicated Zenodo badge in README (by @jfy133)
- #406 - Fix CheckM database always downloading, regardless if CheckM is selected (by @jfy133)
- #419 - Fix bug with busco_clean parameter, where it is always activated (by @prototaxites)
- #426 - Fixed typo in help text for parameters
--host_genome and --host_fasta (by @tillenglert)
- #434 - Fix location of samplesheet for AWS full tests (reported by @Lfulcrum, fix by @jfy133)
- #438 - Fixed version inconsistency between conda and containers for GTDBTK_CLASSIFYWF (by @jfy133)
- #439 - Fix bug in assembly input (by @prototaxites)
- #447 - Remove
default: None from parameter schema (by @drpatelh)
- #449 - Fix results file overwriting in Ancient DNA workflow (reported by @alexhbnr, fix by @jfy133)
- #470 - Fix binning preparation from running even when binning was requested to be skipped (reported by @prototaxites, fix by @jfy133)
- #480 - Improved
-resume reliability through better meta map preservation (reported by @prototaxites, fix by @jfy133)
- #493 - Update
METABAT2 nf-core module so that it reduced the number of unnecessary file moves, enabling virtual filesystems (fix by @adamrtalbot)
- #500 - Fix MaxBin2 bins not being saved in results directly properly (reported by @Perugolate, fix by @jfy133)
| Tool |
Previous version |
New version |
| BCFtools |
1.16 |
1.17 |
| SAMtools |
1.16.1 |
1.17 |
| fastp |
0.23.2 |
0.23.4 |
| MultiQC |
1.14 |
1.15 |
- #461 - Fix full-size AWS test profile paths (by @jfy133)
- #461 - Fix pyDamage results being overwritten (reported by @alexhbnr, fix by @jfy133)
- #458 - Correct the major issue in ancient DNA workflow of binning refinement being performed on uncorrected contigs instead of aDNA consensus recalled contigs (issue #449)
- #451 - Fix results file overwriting in Ancient DNA workflow (reported by @alexhbnr, fix by @jfy133, and integrated by @maxibor in #458 )
- #350 - Adds support for CheckM as alternative bin completeness and QC tool (added by @jfy133 and @skrakau)
- #353 - Added the busco_clean parameter to optionally clean each BUSCO directory after a successful (by @prototaxites)
- #361 - Added the skip_clipping parameter to skip read preprocessing with fastp or adapterremoval. Running the pipeline with skip_clipping, keep_phix and without specifying a host genome or fasta file skips the FASTQC_TRIMMED process (by @prototaxites)
- #365 - Added CONCOCT as an additional (optional) binning tool (by @jfy133)
- #366 - Added CAT_SUMMARISE process and cat_official_taxonomy parameter (by @prototaxites)
- #372 - Allow CAT_DB to take an extracted database as well as a tar.gz file (by @prototaxites).
- #380 - Added support for saving processed reads (clipped, host removed etc.) to results directory (by @jfy133)
- #394 - Added GUNC for additional chimeric bin/contamination QC (added by @jfy133)
- #340,#368,#373 - Update to nf-core 2.7.2
TEMPLATE (by @jfy133, @d4straub, @skrakau)
- #373 - Removed parameter
--enable_conda. Updated local modules to new conda syntax and updated nf-core modules (by @skrakau)
- #385 - CAT also now runs on unbinned contigs as well as binned contigs (added by @jfy133)
- #399 - Removed undocumented BUSCO_PLOT process (previously generated
*.busco_figure.png plots unsuitable for metagenomics) (by @skrakau).
- #416 - Use GTDBTK_CLASSIFYWF nf-core module instead of local module (added by @alxndrdiaz)
- #345 - Bowtie2 mode changed to global alignment for ancient DNA mode (
--very-sensitive mode) to prevent soft clipping at the end of reads when running in local mode. (by @maxibor)
- #349 - Add a warning that pipeline will reset minimum contig size to 1500 specifically MetaBAT2 process, if a user supplies below this threshold. (by @jfy133)
- #352 - Escape the case in the BUSCO module that BUSCO can just detect a root lineage but is not able to find any marker genes (by @alexhbnr)
- #355 - Include error code 21 for retrying with higher memory for SPAdes and hybridSPAdes (by @mglubber)
| Tool |
Previous version |
New version |
| BUSCO |
5.1.0 |
5.4.3 |
| BCFtools |
1.14 |
1.16 |
| Freebayes |
1.3.5 |
1.3.6 |
| SAMtools |
1.15 |
1.16.1 |
- #328 - Fix too many symbolic links issue in local convert_depths module (reported by @ChristophKnapp and fixed by @apeltzer, @jfy133)
- #329 - Each sample now gets it's own result directory for PyDamage analysis and filter (reported and fixed by @maxibor)
- #263 - Restructure binning subworkflow in preparation for aDNA workflow and extended binning
- #247 - Add ancient DNA subworkflow
- #263 - Add MaxBin2 as second contig binning tool
- #285 - Add AdapterRemoval2 as an alternative read trimmer
- #291 - Add DAS Tool for bin refinement
- #319 - Activate pipeline-specific institutional nf-core/configs
- #269,#283,#289,#302 - Update to nf-core 2.4
TEMPLATE
- #286 - Cite our publication instead of the preprint
- #291, #299 - Add extra results folder
GenomeBinning/depths/contigs for [assembler]-[sample/group]-depth.txt.gz, and GenomeBinning/depths/bins for bin_depths_summary.tsv and [assembler]-[binner]-[sample/group]-binDepths.heatmap.png
- #315 - Replace base container for standard shell tools to fix problems with running on Google Cloud
- #290 - Fix caching of binning input
- #305 - Add missing Bowtie2 version for process
BOWTIE2_PHIX_REMOVAL_ALIGN to software_versions.yml
- #307 - Fix retrieval of GTDB-Tk version (note about newer version caused error in
CUSTOM_DUMPSOFTWAREVERSIONS)
- #309 - Fix publishing of BUSCO
busco_downloads/ folder, i.e. publish only when --save_busco_reference is specified
- #321 - Fix parameter processing in
BOWTIE2_REMOVAL_ALIGN (which was erroneously for BOWTIE2_PHIX_REMOVAL_ALIGN)
| Tool |
Previous version |
New version |
| fastp |
0.20.1 |
0.23.2 |
| MultiQC |
1.11 |
1.12 |
- #240 - Add prodigal to predict protein-coding genes for assemblies.
- #241 - Add parameter
--skip_prodigal.
- #244 - Add pipeline preprint information.
- #245 - Add Prokka to annotate binned genomes.
- #249 - Update workflow overview figure.
- #258 - Updated MultiQC 1.9 to 1.11.
- #260 - Updated SPAdes 3.13.1 -> 3.15.3, MEGAHIT 1.2.7 -> 1.2.7
- #256 - Fix
--skip_busco.
- #236 - Fix large assemblies (> 4 billion nucleotides in length).
- #254 - Fix MetaBAT2 error with nextflow version 21.10.x (21.04.03 is the latest functional version for nf-core/mag 2.1.0).
- #255 - Update gtdbtk conda channel.
- #258 - FastP results are now in MultiQC.
- #212, #214 - Add bin abundance estimation based on median sequencing depths of corresponding contigs (results are written to
results/GenomeBinning/bin_depths_summary.tsv and results/GenomeBinning/bin_summary.tsv) #197.
- #214 - Add generation of (clustered) heat maps with bin abundances across samples (using centered log-ratios)
- #217 - Publish genes predicted with Prodigal within BUSCO run (written to
results/GenomeBinning/QC/BUSCO/[assembler]-[bin]_prodigal.gff).
- #218 - Update to nf-core 2.0.1
TEMPLATE (DSL2)
- #226 - Fix handling of
BUSCO output when run in auto lineage selection mode and selected specific lineage is the same as the generic one.
- #179 - Add BUSCO automated lineage selection functionality (new default). The pameter
--busco_auto_lineage_prok can be used to only consider prokaryotes and the parameter --busco_download_path to run BUSCO in offline mode.
- #178 - Add taxonomic bin classification with
GTDB-Tk v1.5.0 (for bins filtered based on BUSCO QC metrics).
- #196 - Add process for CAT database creation as an alternative to using pre-built databases.
- #162 - Switch to DSL2
- #162 - Changed
--input file format from TSV to CSV format, requires header now
- #162 - Update
README.md, docs/usage.md and docs/output.md
- #162 - Update
FastP from version 0.20.0 to 0.20.1
- #162 - Update
Bowtie2 from version 2.3.5 to 2.4.2
- #162 - Update
FastQC from version 0.11.8 to 0.11.9
- #172 - Compressed discarded MetaBAT2 output files
- #176 - Update CAT DB link
- #179 - Update
BUSCO from version 4.1.4 to 5.1.0
- #179 - By default BUSCO now performs automated lineage selection instead of using the bacteria_odb10 lineage as reference. Specific lineage datasets can still be provided via
--busco_reference.
- #178 - Change output file:
results/GenomeBinning/QC/quast_and_busco_summary.tsv -> results/GenomeBinning/bin_summary.tsv, contains GTDB-Tk results as well.
- #191 - Update to nf-core 1.14
TEMPLATE
- #193 - Compress CAT output files #180
- #198 - Requires nextflow version
>= 21.04.0
- #200 - Small changes in GitHub Actions tests
- #203 - Renamed
fastp params and improved description in documentation: --mean_quality -> --fastp_qualified_quality, --trimming_quality -> --fastp_cut_mean_quality
- #175 - Fix bug in retrieving the
--max_unbinned_contigs longest unbinned sequences that are longer than --min_length_unbinned_contigs (split_fasta.py)
- #175 - Improved runtime of
split_fasta.py in METABAT2 process (important for large assemblies, e.g. when computing co-assemblies)
- #194 - Allow different folder structures for Kraken2 databases containing
*.k2d files #187
- #195 - Fix documentation regarding required compression of input FastQ files #160
- #196 - Add process for CAT database creation as solution for problem caused by incompatible
DIAMOND version used for pre-built CAT database and CAT classification #90, #188
- #146 - Add
--coassemble_group parameter to allow group-wise co-assembly
- #146 - Add
--binning_map_mode parameter allowing different mapping strategies to compute co-abundances used for binning (all, group, own)
- #149 - Add two new parameters to allow custom SPAdes and MEGAHIT options (
--spades_options and --megahit_options)
- #141 - Update to nf-core 1.12.1
TEMPLATE
- #143 - Manifest file has to be handed over via
--input parameter now
- #143 - Changed format of manifest input file: requires a '.tsv' suffix and additionally contains group ID
- #143 - TSV
--input file allows now also entries containing only short reads
- #145 - When using TSV input files, uses sample IDs now for
FastQC instead of basenames of original read files. Allows non-unique file basenames.
- #143 - Change parameter:
--manifest -> --input
- #135 - Update to nf-core 1.12
TEMPLATE
- #133 - Fixed processing of
--input parameter #131
- #121 - Add full-size test
- #124 - Add worfklow overview figure to
README
- #123 - Update to new nf-core 1.11
TEMPLATE
- #118 - Fix
seaborn to v0.10.1 to avoid nanoplot error
- #120 - Fix link to CAT database in help message
- #124 - Fix description of
CAT process in output.md
- #35 - Add social preview image
- #49 - Add host read removal with
Bowtie 2 and according custom section to MultiQC
- #49 - Add separate
MultiQC section for FastQC after preprocessing
- #65 - Add
MetaBAT2 RNG seed parameter --metabat_rng_seed and set the default to 1 which ensures reproducible binning results
- #65 - Add parameters
--megahit_fix_cpu_1, --spades_fix_cpus and --spadeshybrid_fix_cpus to ensure reproducible results from assembly tools
- #66 - Export
depth.txt.gz into result folder
- #67 - Compress assembly files
- #82 - Add
nextflow_schema.json
- #104 - Add parameter
--save_busco_reference
- #56 - Update
MetaBAT2 from v2.13 to v2.15
- #46 - Update
MultiQC from v1.7 to v1.9
- #88 - Update to new nf-core 1.10.2
TEMPLATE
- #88 -
--reads is now removed, use --input instead
- #101 - Prevented PhiX alignments from being stored in work directory #97
- #104, #111 - Update
BUSCO from v3.0.2 to v4.1.4
- #29 - Fix
MetaBAT2 binning discards unbinned contigs #27
- #31, #36, #76, #107 - Fix links in README
- #47 - Fix missing
MultiQC when --skip_quast or --skip_busco was specified
- #49, #89 - Added missing parameters to summary
- #50 - Fix missing channels when
--keep_phix is specified
- #54 - Updated links to
minikraken db
- #54 - Fixed
Kraken2 dp preparation: allow different names for compressed archive file and contained folder as for some minikraken dbs
- #55 - Fixed channel joining for multiple samples causing
MetaBAT2 error #32
- #57 - Fix number of threads used by
MetaBAT2 program jgi_summarize_bam_contig_depths
- #70 - Fix
SPAdes memory conversion issue #61
- #71 - No more ignoring errors in
SPAdes assembly
- #72 - No more ignoring of
BUSCO errors
- #73, #75 - Improved output documentation
- #96 - Fix missing bin names in
MultiQC BUSCO section #78
- #104 - Fix
BUSCO errors causing missing summary output #77
- #29 - Change depreciated parameters:
--singleEnd -> --single_end, --igenomesIgnore -> --igenomes_ignore
Initial release of nf-core/mag, created with the nf-core template.
- short and long reads QC (fastp, porechop, filtlong, fastqc)
- Lambda and PhiX detection and filtering (bowtie2, nanolyse)
- Taxonomic classification of reads (centrifuge, kraken2)
- Short read and hybrid assembly (megahit, metaspades)
- metagenome binning (metabat2)
- QC of bins (busco, quast)
- annotation (cat/bat)