Pandora is a tool for bacterial genome analysis using a pangenome reference graph (PanRG). It allows gene presence/absence detection and genotyping of SNPs, indels and longer variants in one or a number of samples. Pandora works with Illumina or Nanopore data. Core ideas behind the method are:
- new genomes look like recombinants (plus mutations) of things seen before
- we should be analysing nucleotide-level variation everywhere, not just in core genes
- arbitrary single reference genomes are unnatural and limit comparisons of diverse sets of genomes
The pangenome reference graph (PanRG) is a collection of 'floating' local graphs, each representing some orthologous region of interest (e.g. genes, mobile elements or intergenic regions). See https://github.com/rmcolq/make_prg for a pipeline which can construct these PRGs from a set of aligned sequence files.
Pandora can do the following for a single sample (read dataset):
- Output inferred mosaic of reference sequences for loci (eg genes) from the PanRG which are present
- Output a VCF showing the variation found within these loci, with respect to any reference path in the PRG.
Soon, in a galaxy not so far away, it will allow:
- discovery of new variation not in the PRG
For a collection of samples, it can:
- Output a matrix showing inferred copy-number of each locus in each sample genome
- Output a multisample pangenome VCF showing how including genotype calls for each sample in each of the loci
- Output one VCF per orthologous-chunk, showing how samples which contained this chunk differed in their gene sequence. Variation is shown with respect to the most informative recombinant path in the PRG.
Warning - this code is still in development.
Index PanRG file:
pandora index -t 8 <panrg.fa>
Compare first 30X of each Illumina sample to get pangenome matrix and VCF
pandora compare -p <panrg.fa> -r <read_index.tab> --genotype --illumina --max_covg 30
Map Nanopore reads from a single sample to get approximate sequence for genes present
pandora map -p <panrg.fa> -r <reads.fq>
We highly recommend that you download and use the singularity container:
singularity pull shub://rmcolq/pandora:pandora
or download direcly as you run:
singularity exec shub://rmcolq/pandora:pandora pandora <command>
This is not recommended because the required zlib and boost system installs do not always play nicely. If you want to take the risk:
- Requires a Unix or Mac OS.
- Requires a system install of
zlib. If this is not already installed, this tutorial is helpful or try the following.
wget http://www.zlib.net/zlib-1.2.11.tar.gz -O - | tar xzf -
cd zlib-1.2.11
./configure [--prefix=/prefix/path]
make
make install
- Requires a system installation of
boostcontaining thesystem,filesystem,log(which also depends onthreadanddate_time) andiostreamslibraries. If not already installed use the following or look at this guide.
wget https://sourceforge.net/projects/boost/files/boost/1.62.0/boost_1_62_0.tar.gz -O - | tar xzf -
cd boost_1_62_0
./bootstrap.sh [--prefix=/prefix/path] --with-libraries=system,filesystem,iostreams,log,thread,date_time
./b2 install
- Download and install
pandoraas follows:
git clone --single-branch https://github.com/rmcolq/pandora.git --recursive
cd pandora
mkdir -p build
cd build
cmake ..
make
ctest -VV
Pandora assumes you have already constructed a fasta-like file of graphs, one entry for each gene/ genome region of interest. If you haven't, you will need a multiple sequence alignment for each graph. Precompiled collections of MSA representing othologous gene clusters for a number of species can be downloaded from here and converted to graphs using the pipeline from here.
Takes a fasta-like file of PanRG sequences and constructs an index, and a directory of gfa files to be used by pandora map or pandora compare. These are output in the same directory as the PanRG file.
Usage: pandora index [options] <PanRG>
Options:
-h,--help Show this help message
-w W Window size for (w,k)-minimizers, default 14
-k K K-mer size for (w,k)-minimizers, default 15
-t T Number of concurrent threads, default 1
The index stores (w,k)-minimizers for each PanRG path found. These parameters can be specified, but default to w=14, k=15.
This takes a fasta/q of Nanopore or Illumina reads and compares to the index. It infers which of the PanRG genes/elements is present, and for those that are present it outputs the inferred sequence and a genotyped VCF.
Usage: pandora map -p PanRG_FILE -r READ_FILE -o OUTDIR <option(s)>
Options:
-h,--help Show this help message
-p,--prg_file PanRG_FILE Specify a fasta-style PanRG file
-r,--read_file READ_FILE Specify a file of reads in fasta/q format
-o,--outdir OUTDIR Specify directory of output
-w W Window size for (w,k)-minimizers, must be <=k, default 14
-k K K-mer size for (w,k)-minimizers, default 15
-m,--max_diff INT Maximum distance between consecutive hits within a cluster, default 250 bps
-e,--error_rate FLOAT Estimated error rate for reads, default 0.11/0.001 for Nanopore/Illumina
-c,--min_cluster_size INT Minimum number of hits in a cluster to consider a locus present, default 10
--genome_size NUM_BP Estimated length of genome, used for coverage estimation, default 5000000
--vcf_refs REF_FASTA A fasta file with an entry for each loci in the PanRG in order, giving
reference sequence to be used as VCF ref. Must have a perfect match to a
path in the graph and the same name as the locus in the graph.
--illumina Data is from Illumina, not Nanopore, so is shorter with low error rate
--bin Use binomial model for kmer coverages, default is negative binomial
--max_covg INT Maximum average coverage from reads to accept, default first 300
--genotype Output a genotyped VCF
--discover Add denovo discovery
--denovo_kmer_size INT Kmer size to use for denovo discovery, default 11
--log_level LEVEL Verbosity for logging, use "debug" for more output
This takes Nanopore or Illumina read fasta/q for a number of samples, mapping each to the index. It infers which of the PanRG genes/elements is present in each sample, and outputs a presence/absence pangenome matrix, the inferred sequences for each sample and a genotyped multisample pangenome VCF.
Usage: pandora compare -p PanRG_FILE -r READ_INDEX -o OUTDIR <option(s)>
Options:
-h,--help Show this help message
-p,--prg_file PanRG_FILE Specify a fasta-style PanRG file
-r,--read_index READ_INDEX Specify a tab delimited file with a line per sample, detailing sample id
and read fasta/q
-o,--outdir OUTDIR Specify directory of output
-w W Window size for (w,k)-minimizers, must be <=k, default 14
-k K K-mer size for (w,k)-minimizers, default 15
-m,--max_diff INT Maximum distance between consecutive hits within a cluster, default 250 bps
-e,--error_rate FLOAT Estimated error rate for reads, default 0.11/0.001 for Nanopore/Illumina
-c,--min_cluster_size INT Minimum number of hits in a cluster to consider a locus present, default 10
--genome_size NUM_BP Estimated length of genome, used for coverage estimation, default 5000000
--vcf_refs REF_FASTA A fasta file with an entry for each loci in the PanRG in order, giving
reference sequence to be used as VCF ref. Must have a perfect match to a
path in the graph and the same name as the locus in the graph.
--illumina Data is from Illumina, not Nanopore, so is shorter with low error rate
--bin Use binomial model for kmer coverages, default is negative binomial
--max_covg INT Maximum average coverage from reads to accept, default first 300
--genotype Output a genotyped VCF
--log_level LEVEL Verbosity for logging, use "debug" for more output