deepvariant

DeepVariant is an analysis pipeline that uses a deep neural network to call genetic variants from next-generation DNA sequencing data

variant callingmachine learningneural network

https://github.com/nf-core/modules/[...]/subworkflows/nf-core/deepvariant

Description

DeepVariant is an analysis pipeline that uses a deep neural network to call genetic variants from next-generation DNA sequencing data

Input

name

description

pattern

`ch_input`

Input aligned reads in bam or cram format, with index, and optional intervals BED file
Structure: [ val(meta), path(bam_or_cram), path(bai_or_crai), path(intervals_bed) ]

`ch_fasta`

Reference genome
Structure: [ val(meta2), path(fasta) ]

`ch_fai`

Reference genome index in fai format
Structure: [ val(meta3), path(fai) ]

`ch_gzi`

Reference genome index in gzi format (either gzi or fai should be used)
Structure: [ val(meta4), val(gzi) ]

`ch_par_bed`

bed file of pseudoautosomal regions (optional)
Structure: [ val(meta5), val(par_bed) ]

*.bed

`with_phasing`

Whether to enable haplotype phasing of the variant calls

Output

name

description

pattern

`vcf`

Variant calls
Structure: [ val(meta), path(vcf) ]

*.vcf.gz

`vcf_index`

Index for variant call file
Structure: [ val(meta), path(vcf_index) ]

*.tbi

`gvcf`

Variant call file with genomic coverage information
Structure: [ val(meta), path(gvcf) ]

*.g.vcf.gz

`gvcf_index`

Index for the GVCF.
Structure: [ val(meta), path(gvcf_index) ]