Accurate genotyping of known variants is a critical for the analysis of whole-genome sequencing data. Paragraph aims to facilitate this by providing an accurate genotyper for Structural Variations with short-read data.
Please reference Paragraph using:
Genotyping data in this paper can be found at paper-data/download-instructions.txt
For details of population genotyping, please also refer to:
Please check doc/Installation.md for system requirements and installation instructions.
After installation, run multigrmpy.py script from the build/bin directory on an example dataset as follows:
python3 bin/multigrmpy.py -i share/test-data/round-trip-genotyping/candidates.vcf \
-m share/test-data/round-trip-genotyping/samples.txt \
-r share/test-data/round-trip-genotyping/dummy.fa \
-o test \This runs a simple genotyping example for two test samples.
chr1The output folder test then contains gzipped json for final genotypes:
$ tree testtest
├── grmpy.log # main workflow log file
├── genotypes.vcf.gz # Output VCF with individual genotypes
├── genotypes.json.gz # More detailed output than genotypes.vcf.gz
├── variants.vcf.gz # The input VCF with unique ID from Paragraph
└── variants.json.gz # The converted graphs from input VCF (no genotypes)If successful, the last 3 lines of genotypes.vcf.gz will the same as in expected file.
paraGRAPH will independently genotype each entry of the input VCF. You can use either indel-style representation (full REF and ALT allele sequence in 4th and 5th columns) or symbolic alleles, as long as they meet the format requirement of VCF 4.0+.
Currently we support 4 symbolic alleles:
<DEL> for deletion
<INS> for insertion
<DUP> for duplication
<INV> for inversion
Must be tab-deliemited.
Required columns:
Optional columns:
On a 30x HiSeqX sample, Paragraph typically takes 1-2 seconds to genotype a simple SV in confident regions.
If the SV is in a low-complexity region with abnormal read pileups, the running time could vary.
For efficiency, it is recommended to manually set the "-M" option (maximum allowed read count for a variant) to skip these high-depth regions. We recommend "-M" as 20 times of your mean sample depth.
To efficiently genotype SVs across a population, we recommend doing single-sample mode as follows:
multigrmpy.py for each manifest. Be sure to set "-M" option for each sample according to its depth.genotypes.vcf.gz to create a big VCF of all samples. You can use either bcftools merge or your custom script.For more complicated events (e.g. genotype a deletion together with its nearby SNP), you can provide a custimized JSON to paraGRAPH:
Please follow the pattern in example JSON and make sure all required keys are provided. Here is a visualization of this sample graph.
To obtain graph alignments for this graph (including all reads), run:
bin/paragraph -b <input BAM> \
-r <reference fasta> \
-g <input graph JSON> \
-o <output JSON path> \
-E 1To obtain the algnment summary, genotypes of each breakpoint, and the whole graph, run:
bin/grmpy -m <input manifest> \
-r <reference fasta> \
-i <input graph JSON> \
-o <output JSON path> \
-E 1If you have multiple events listed in the input JSON, multigrmpy.py can help you to run multiple grmpy jobs together.
Please check github wiki for common usage questions and errors.
More information about all tools we provide in this package can be found in doc/graph-tools.md.
In doc/graph-models.md we describe the graph and genotyping models we implement.
Some developer documentation about our code analysis and testing process can be found in doc/linting-and-testing.md.
Procedures for read level alignment validation doc/validation-with-simulated-reads.md.
How we count reads for variants and paths doc/graph-counting.md.
Documentation of genotyping model parameters doc/genotyping-parameters.md.
Doc/graphs-ashg-2017.pdf contains the poster about this method we showed at ASHG 2017
The LICENSE file contains information about libraries and other tools we use, and license information for these.