CoreGenomePrimers (CGP)

Summary

Pangenome-guided primer design for clade-specific primer generation

Overview of the method:

• You provide a set of genomes for the target clade
  • e.g., all Streptomyces griseus genomes
• All coding & rRNA gene sequences for all genomes are clustered in order to find "core" genes
  • "core" = found in X fraction of members (set in the config)
    • "core" also means single-copy for CDS genes, by default (set in the config)
• For each core gene cluster:
  • Sequences are aligned to produce a multiple sequence alignment (MSA)
  • Primers are designed from the MSA via primer3
  • Non-specific primers are filtered based on blastn hits to non-target gene clusters
  • Non-specific primers are filtered based on blastn hits to non-target taxa
    • e.g., everything in NCBI-nt except for the target clade
• Supplemental info
  • BLAST-based gene annotations of each core gene cluster
  • in silco amplicons generated by seqkit amplicon

Pipeline overview

Rule graph:

Workflow

• Input
  • Table of target genomes and related info
    • See Pectobacterium.tsv for an example
  • Config file for setting pipeline parameters
    • See config.yaml
• Core gene cluster identification
  • Genes called/annoated with prokka
  • Genes clustered with mmseqs cluster (or linclust)
  • Core genes defined by default as present in all members (and single copy) in each genome
• Primer design
  • Core genes aligned with mafft
  • Primers generated from the MSA via primer3
    • Method:
      • A majority-rule consensus sequence is generated from the MSA
        • Non-majority positions are assigned "N"
      • primer3 is used to design primers based on the consensus sequence
        • Only primers that are relatively conserved regions are generated
        • primer3 can constrain primer creation based on many criteria, such as:
          • amplicon length
          • melting temperature (Tm)
          • G+C content
      • The consensus sequence primers are not degenerate (primer3 cannot do that)
      • To add degeneracies and thus increase primer-set specificity to the gene region:
        • The sub-sequences of the MSA where the raw primers hit are extracted
    • A degenerate primer sequence is produced from all sub-sequences
  • The raw, degenerate primers are filtered by:
    • Total degeneracy
    • Degeneracy at the 3' end (higher degeneracy = less primer annealing)
  • Primers assessed for cross-hybridization to non-target genes in each target genome
    • blastn of primers against all gene clusters
    • blast hits filtered hit percent-identity, hit length, and amplicon size
  • Primers assessed for non-target hits to non-target taxa
    • The target clade is designated by the Taxid column in the input table.
      • All members in the database with that taxid will be excluded from the blast search.
    • blastn of primers against NCBI-nt (by default)
    • blast hits filtered hit percent-identity, hit length, and amplicon size
  • Final primer info is compiled in the output directory
• Supplemental info
  • Core genes BLAST'ed against nr (by default) while excluding the target clade in order to get most closely related non-target genes
    • This information can help prioritize primers that target core genes that are most unique to the target clade

Install

Snakemake

The pipeline utilizes snakemake. We recommend that you install it via conda or mamba.

Pipeline

git clone --recurse-submodules git@github.com:leylabmpi/CoreGenomePrimers.git

Databases

BLAST (nucl & prot)

Use update_blastdb.pl or another method

BLAST (rRNA)

See the NCBI rRNA databases

You can also download the SSU and LSU databases from ftp:/ftp/ebio/projects/CoreGenomePrimers/

Gene names

You can download the gene names pkl file from ftp:/ftp/ebio/projects/CoreGenomePrimers/

Taxonomy

See taxonkit

You can also download the taxonomy files from ftp:/ftp/ebio/projects/CoreGenomePrimers/

Notes

Make sure to update the file paths to the databases in the config.yaml

Usage instructions

For general instuctions on using snakemake, see the snakemake docs.

Input

Samples file

Set in the config.yaml file via the samples_file: parameter

See Pectobacterium.tsv for an example

A tab-delimited file with the following columns:

• Taxon
  • Taxon name for the genome
• Fasta
  • Path to the fasta file for the genome assembly
    • The assembly can be incomplete
    • Contig naming does not matter
• Taxid
  • NCBI taxonomy ID for the genome
    • This will be used to exclude this clade from BLAST searchers against non-target clades
    • If a non-species-level taxid is provided, taxonkit is used to find the species-level taxid(s)
      • BLAST -negative_taxidlist can only use species-level taxids
• Domain
  • Taxonomic domain of the genome (used by prokka)

Key config params

Example config file: config.yaml

The default parameters in config.yaml are set for qPCR design (but no internal oligo).

The following are parameters that you most likely would want to change for your own needs.

• core_genes:
  • --perc-genomes-(cds|rrna)
    • The % of target genomes that must contain the gene cluster
  • --copies-per-genome-(cds|rrna)
    • The max number of copies in a genome in order to be considered "core"
  • --max-clusters
    • The max number of core genes that will be used for downstream analyses
• blast_nontarget
  • parameters for blast(n|x) of the core genes vs NCBI-nr (by default) in order to get non-target clade hits
• align
  • Parameters passed to mafft
• `primer3:
  • number:
    • --num-primers
      • Max number of raw primers generated
  • size:
    • Optimum/minimum/maximum primer size
  • `product:
    • Optimum/minimum/maximum primer amplicon size
  • Tm:
    • Optimum/minimum/maximum primer melting temperature
    • --max-tm-diff
      • The max allowable difference in Tm between oligos in the primer set
  • GC:
    • Optimum/minimum/maximum primer G+C content
  • degeneracy:
    • --max-degeneracy
      • Max degeneracy of the oligo
    • --max-degeneracy-3prime
      • Max degeneracy at the 3' end of either primer
    • --window-3prime
      • The size (bp) of the 3' region to consider for --max-degeneracy-3prime
• blastn:
  • -max_target_seqs
    • increase for extra stringency when checking for non-target hits
  • -perc_identity
    • decrease for extra stringency when checking for non-target hits
• blastn_filter:
  • --perc-len
    • What % length of blast hit (hit_len / query_len * 100) is considered a legitimate hit?
  • --(min|max)-amplicon-len
    • fwd-rev primer hits generated amplicons outside of this size range will not be considered legitimate non-target hits

Output

Notes on terminology

• degenerate oligo = oligo nucleotide sequence with degeneracies
  • Example: ATACGTAY
• expanded oligo = all non-degenerate sequences encoded by the degenerate sequence
  • Example: ATACGTAY = (ATACGTAC, ATACGTAT)

Description

• $OUTDIR/core_clusters_info.tsv
  • General metadata on all core gene clusters
• $OUTDIR/nontarget/*_blast*.tsv
  • Formatted blast hits for a representative of each core gene cluster
  • Just blast hits to non-target taxa (default by Taxid column in the input table)
• $OUTDIR/clusters/
  • Core gene cluster fasta, alignment, and metadata files
• $OUTDIR/primers_final/
  • fasta files & metadata tables for each primer
  • _degen.fna => degenenerate primer sequences
  • _expand.fna => degeneracies "expanded" to non-ambiguous characters
• $OUTDIR/primers_final_info.tsv
  • Summary info of all primers that passes all filtered
  • See below for column descriptions
• $OUTDIR/amplicons.tsv
  • in silico amplicons generated by seqkit seq
    • primer set search to target gene cluster

Column descriptions

$OUTDIR/core_clusters_info.tsv

• gene_type
  • CDS or rRNA gene
• cluster_id
  • Numeric ID of the gene cluster
• seq_uuid
  • Universal unique ID given to the gene sequence
• seq_orig_name
  • Original gene annotation (replaced by the UUID for use in the pipeline)
• contig_id
  • Name of the contig that contains the gene
• taxon
  • Name of the taxon that contains the gene
• start
  • Start position (bp) on the contig containing the gene
• end
  • End position (bp) on the contig containing the gene
• score
  • Prodigal annotation score
• strand
  • Sequence strand that contains the contig
• cluster_name
  • Name of the cluster (the seq_uuid of the cluster representative sequence)

$OUTDIR/primers_final_info.tsv

• gene_type
  • CDS or rRNA gene
• cluster_id
  • The ID of the core gene cluster
• primer_set
  • The ID of the primer set
  • NOTE: The primer sets are numbered per-cluster, so IDs repeat among gene clusters
• amplicon_size_consensus
  • Amplicon size (bp) of the gene cluster alignment consensus sequence
• amplicon_size_avg
  • Average size (bp) of the putative amplicons
  • This may differ from amplicon_size_consensus due to gaps in some of the gene sequences
• amplicon_size_sd
  • Standard deviation of sizes (bp) of the putative amplicons
• primer_id
  • primer_set + [f,r,i]
    • f = forward primer
    • r = reverse primer
    • i = internal oligo
• primer_type
  • PRIMER_RIGHT = forward primer
  • PRIMER_LEFT = reverse primer
  • PRIMER_INTERNAL = interal oligo
• sequence
  • Oligo sequence
• length
  • Oligo length
• degeneracy
  • Oligo degeneracy
• degeneracy_3prime
  • Degeneracy only at the 3' end of the oligo
  • The 3' end is determined via --window-3prime
• position_start
  • Start position of the oligo on the gene cluster MSA consensus sequence
• position_end
  • End position of the oligo on the gene cluster MSA consensus sequence
• Tm_avg
  • Average melting temperature for all expanded oligo sequences
• Tm_sd
  • Standard deviation of melting temperatures for all expanded oligo sequences
• GC_avg
  • Average G+C content for all expanded oligo sequences
• GC_sd
  • Standard deviation of G+C contents for all expanded oligo sequences
• hairpin_avg
  • Average oligo hairpin melting temperature for all expanded oligo sequences
• hairpin_sd
  • Standard deviation of oligo hairpin melting temperatures for all expanded oligo sequences
• homodimer_avg
  • Average oligo homodimer melting temperature for all expanded oligo sequences
• homodimer_sd
  • Standard deviation of oligo homodimer melting temperatures for all expanded oligo sequences

$OUTDIR/nontarget/*_blast*.tsv

• cluster_id
  • Core gene cluster ID number
• query
  • The seq_uuid of the core gene cluster representative sequence
• subject
  • The accession of the BLAST subject
• subject_name
  • The gene annotation (and taxonomy) of the subject
• Other columns
  • See the BLAST documentation for column descriptions

$OUTDIR/amplicons.tsv

• gene_type
  • CDS or rRNA gene
• cluster_id
  • Core gene cluster ID number
• gene_id
  • The seq_uuid of the gene sequence in the cluster
• start
  • amplicon start position
• end
  • amplicon end position
• primer_set
  • The ID of the primer set
  • NOTE: The primer sets are numbered per-cluster, so IDs repeat among gene clusters
• score
  • ignore
• strand
  • strand of amplicon
• amplicon
  • amplicon sequence

Misc

Summarizing primer design/filtering logs

See ./utils/primer_log_summary.py