Human variation workflow

SNV, SV and CNV calling, modified base calling, and STR genotyping of human samples.

Introduction

This repository contains a nextflow workflow for analysing variation in human genomic data. Specifically this workflow can perform the following:

• diploid variant calling
• structural variant calling
• analysis of modified base calls
• copy number variant calling
• short tandem repeat (STR) expansion genotyping

The wf-human-variation workflow consolidates the small variant calling from the previous wf-human-snp, structural variant calling from wf-human-sv, CNV calling from wf-cnv (all of which are now deprecated), as well as performing STR expansion genotyping. This pipeline performs the steps of the four pipelines simultaneously and the results are generated and output in the same way as they would have been had the pipelines been run separately.

Please note, the tools embedded in individual sub-workflows within wf-human-variation are intended for use with 20x whole-genome Oxford Nanopore Technologies sequencing data (with the exception of QDNAseq, please see this section for more information). Usage outside of this (e.g. with adaptive sampling data, or using lower coverage inputs) may cause the workflow to terminate with an error, or produce unexpected results.

Compute requirements

Recommended requirements:

• CPUs = 32
• Memory = 128GB

Minimum requirements:

• CPUs = 16
• Memory = 32GB

Approximate run time: Variable depending on whether it is targeted sequencing or whole genome sequencing, as well as coverage and the individual analyses requested. For instance, a 90X human sample run (options: --snp --sv --mod --str --cnv --phased --sex XY) takes less than 8h with recommended resources.

ARM processor support: False

Install and run

These are instructions to install and run the workflow on command line. You can also access the workflow via the EPI2ME Desktop application.

The workflow uses Nextflow to manage compute and software resources, therefore Nextflow will need to be installed before attempting to run the workflow.

The workflow can currently be run using either Docker or Singularity to provide isolation of the required software. Both methods are automated out-of-the-box provided either Docker or Singularity is installed. This is controlled by the -profile parameter as exemplified below.

It is not required to clone or download the git repository in order to run the workflow. More information on running EPI2ME workflows can be found on our website.

The following command can be used to obtain the workflow. This will pull the repository in to the assets folder of Nextflow and provide a list of all parameters available for the workflow as well as an example command:

nextflow run epi2me-labs/wf-human-variation --help

To update a workflow to the latest version on the command line use the following command:

nextflow pull epi2me-labs/wf-human-variation

A demo dataset is provided for testing of the workflow. It can be downloaded and unpacked using the following commands:

wget https://ont-exd-int-s3-euwst1-epi2me-labs.s3.amazonaws.com/wf-human-variation/hg002%2bmods.v1/wf-human-variation-demo.tar.gz
tar -xzvf wf-human-variation-demo.tar.gz

The workflow can then be run with the downloaded demo data using:

nextflow run epi2me-labs/wf-human-variation \
    --bam 'wf-human-variation-demo/demo.bam' \
    --ref 'wf-human-variation-demo/demo.fasta' \
    --bed 'wf-human-variation-demo/demo.bed' \
    --sample_name 'DEMO' \
    --snp \
    --sv \
    --mod \
    --phased \
    -profile standard

For further information about running a workflow on the command line see https://labs.epi2me.io/wfquickstart/

Related protocols

This workflow is designed to take input sequences that have been produced from Oxford Nanopore Technologies devices.

Find related protocols in the Nanopore community.

Input example

The --bam input parameter for this workflow accepts a path to a single BAM file, or a folder containing multiple BAM files for the sample. A sample name can be supplied with --sample.

(i)                     (ii)    
input_reads.bam     ─── input_directory
                        ├── reads0.bam
                        └── reads1.bam

Input parameters

Workflow Options

Main options

Copy number variant calling options

Modified base calling options

Short tandem repeat expansion genotyping options

Advanced Options

Multiprocessing Options

Outputs

Output files may be aggregated including information for all samples or provided per sample. Per-sample files will be prefixed with respective aliases and represented below as {{ alias }}.

Pipeline overview

The workflow is composed of 6 distinct subworkflows, each enabled by a command line option:

• SNP calling: --snp
• SV calling: --sv
• Analysis of modified bases: --mod
• CNV calling (Spectre): --cnv
• CNV calling (QDNAseq): --cnv --use_qdnaseq
• STR genotyping: --str

Subworkflows where the relevant option is omitted will not be run.

1. Input and data preparation

The workflow relies on two primary input files:

1. A reference genome in FASTA format
2. Sequencing data for the sample in the form of a single BAM file or a folder of BAM files, either aligned or unaligned.

When analysing human data, we recommend using human_g1k_v37.fasta.gz or GCA_000001405.15_GRCh38_no_alt_analysis_set.fna.gz. For more information see this blog post which outlines potential pitfalls with the various flavours of human references.

The input BAM file can be generated using the wf-basecalling workflow, which is up to date with the current dorado releases and models.

2. Data QC and pre-processing

The workflow starts by performing multiple checks of the input BAM file, as well as computing:

1. depth of sequencing with mosdepth;
2. read alignment statistics with fastcat.

After computing the coverage, the workflow will check that the input BAM file has a depth greater than --bam_min_coverage. In case the user specify --bam_min_coverage 0, the check will be skipped and the workflow will proceed directly to the downstream analyses. Some components work better withing certain ranges of coverage, and the user might achieve better results by providing a target coverage to downsample to. The user can set --downsample_coverage true to enable the downsampling of the reads, and --downsample_coverage_target {{ X }} to specify the target coverage (default: 60x).

An optional gene summary may also be generated by specifying --output_gene_summary. A 4-column BED file is required for this, where the fourth column contains the gene label.

3. Small variant calling with Clair3

The workflow implements a deconstructed version of Clair3 (v1.0.4) to call germline variants. The workflow will select an appropriate Clair3 model by detecting the basecall model from the input data. If the input data does not have the required information to determine the basecall model, the workflow will require the basecall model to be provided explicitly with the --override_basecall_cfg option. This workflow takes advantage of the parallel nature of Nextflow, providing optimal efficiency in high-performance, distributed systems. The workflow will automatically call small variants (SNPs and indels), collect statistics, annotate them with SnpEff (and additionally for SNPs, ClinVar details), and create a report summarising the findings.

If desired, the workflow can perform phasing of structural variants by using the --phased option. This will lead the workflow to use whatshap to perform phasing of the variants, with the option to use longphase instead by setting --use_longphase true. The phasing will also generate a GFF file with the annotation of the phase blocks, enabling the visualisation of these blocks in genome browsers.

4. Structural variant (SV) calling with Sniffles2

The workflow allows for calling of SVs using long-read sequencing data with Sniffles2. The workflow will perform SV calling, filtering and generation of a report. The SV workflow can accept a tandem repeat annotations BED file to improve calling in repetitive regions --- see the sniffles documentation for more information. The workflow will attempt to select an appropriate hg19 or hg38 TR BED but you can override the BED file used with the --tr_bed parameter. SVs can be phased using --phased. However, this will cause the workflow to run SNP analysis, as SV phasing relies on the haplotagged reads generated in this stage.

5. Modified base calling with modkit

Modified base calling can be performed by specifying --mod. The workflow will call modified bases using modkit. The workflow will automatically check whether the files contain the appropriate MM/ML tags, required for running modkit pileup. If the tags are not found, the workflow will not run the individual analysis, but will still run the other subworkflows requested by the user. The default behaviour of the workflow is to run modkit with the --cpg --combine-strands options set. It is possible to report strand-aware modifications by providing --force_strand, which will trigger modkit to run in default mode. The resulting bedMethyl will include modifications for each site on each strand separately. The modkit run can be fully customized by providing --modkit_args. This will override any preset, and allow full control over the run of modkit. Haplotype-resolved aggregated counts of modified bases can be obtained with the --phased option. This will generate three distinct BEDMethyl files with the naming pattern {{ alias }}.wf_mods.{{ haplotype }}.bedmethyl.gz, where haplotype can be 1, 2 or ungrouped.

6a. Copy number variants (CNV) calling with Spectre

CNV calling is performed using an ONT implementation of Spectre using the --cnv flag. Spectre is the default CNV caller in the workflow and is compatible with genome builds hg19/GRCh37 or hg38/GRCh38. The output of this workflow is a VCF of CNV calls annotated with SnpEff. Advanced users may wish to override default parameters using --spectre_args, e.g. --spectre_args "--min-cnv-len 100000". We don’t recommend setting --min-cnv-len to less than N50 * 5 to avoid false positive calls.

6b. Copy number variants (CNV) calling with QDNASeq

CNV calling may alternatively be performed using QDNAseq, using --cnv --use_qdnaseq. This workflow is compatible with genome builds hg19/GRCh37 or hg38/GRCh38, and is recommended for shallow WGS or adaptive sampling data. In addition to the VCF of CNV calls, the workflow emits QDNAseq-generated plots and BED files of both raw read counts per bin and corrected, normalised, and smoothed read counts per bin. Please note that QDNAseq was the default CNV caller until version 1.11.0 of the workflow, and the additional --use_qdnaseq flag is now required to use it.

7. Short tandem repeat (STR) genotyping with Straglr

STR genotyping is performed using a fork of straglr. This workflow is compatible with genome build hg38/GRCh38. The number of calls for repeats on chrX is dependent on the sample’s genetic sex which should be provided if known with --sex XX or --sex XY. If --sex is not specified, the workflow will attempt to infer the genetic sex from coverage of the allosomes, falling back to XX if a determination is unclear. Please be aware that incorrect sex assignment will result in the wrong number of calls for all repeats on chrX. In addition to a gzipped VCF file containing STRs found in the dataset, the workflow emits a TSV straglr output containing reads spanning STRs, and a haplotagged BAM.

8. Phasing variants

Variant phasing is switched on simply using the --phased option. By default, the workflow uses whatshap to perform phasing of the variants, with the option to use longphase instead by setting --use_longphase true. The workflow will automatically turn on the necessary phasing processes based on the selected subworkflows. The behaviour of the phasing is summarised in the below table:

Using --GVCF together with --phased will generate a phased GVCF, created by reflecting the phased genotype and the phase set annotation in the VCF file. This operation is performed using bcftools annotate, targeting the GT and PS fields.

Running the phasing is a compute intensive process. Running the workflow in phasing mode doubles the runtime, and significantly increases the storage requirements to the order of terabytes.

9. Variant annotation

Annotation will be performed automatically by the SNP and SV subworkflows, and can be disabled by the user with --annotation false. The workflow will annotate the variants using SnpEff, and currently only support the human hg19 and hg38 genomes. Additionally, the workflow will add the ClinVar annotations for the SNP variants.

Running the workflow on non-human samples will require this option to be disabled. For more detail, see Section 10 below.

10. Genome compatibility and running the workflow on non-human genomes

Some of the sub-workflows in wf-human-variation are restricted to certain genome builds, which means they will not be executable on non-human genomes or human genome builds outside hg19/GRCh37 and hg38/GRCh38. The following table summarises which subworkflows and options are available (or required) for a desired input genome:

* As noted above, annotation is performed by default but can be switched off for hg19/GRCh37 and hg38/GRCh38.

Troubleshooting

• Annotations for --snp and --sv are generated using SnpEff. For --snp, additional ClinVar annotations are displayed in the report where available (please note, the report will not display any variants classified as ‘Benign’ or ‘Likely benign’, however these variants will be present in the output VCF).
• Aggregation of modified calls with --mod requires data to be basecalled with a model that includes base modifications, providing the MM and ML BAM tags
• CRAM files generated within the workflow cannot be read without the corresponding reference
• The STR workflow performs genotyping of specific repeats, which can be found here.
• By default, SNPs and SVs will be called on chromosomes 1-22, X, Y and MT; to call sites on other sequences enable the --include_all_ctgs option.
• While designed to work on human genomes, the workflow can be run on non-human species by setting --cnv false --str false --annotation false --include_all_ctgs true.
• Ensure that the provided reference and BED files use the same chromosome coding (for example, that they both have the chr prefix, or they both to not have it).
• If unaligned reads were provided, the workflow will output a CRAM file (or BAM if the user runs the --cnv option) containing the alignments used to make the downstream variant calls.
• Renaming, moving or deleting the input BAM, reference genome or the output directory from the location provided at runtime will cause IGV not to load.
• The workflow expects either an uncompressed or bgzip-compressed reference. If the user provides a reference compressed not with bgzip, the workflow will run to completion, but won’t be able to generate the necessary indexes to visualize the outputs in IGV.

FAQ’s

If your question is not answered here, please report any issues or suggestions on the github issues page or start a discussion on the community.

• The number of SNVs and indels in the report do not sum up to the number of record, is that normal? - Yes; this can be due to some multiallelic sites carrying a mixture of SNV and indel alleles.

Related blog posts

• Importing third-party workflows into EPI2ME Labs
• blogpost and CNV workflow documentation for more information on running the copy number calling subworkflow.
• blogpost on how to run a targeted BRCA Gene Analysis with wf-human-variation
• blogpost on the generation of four Genome in a Bottle samples processed with wf-human-variation

See the EPI2ME website for lots of other resources and blog posts.