What GATK Is and How to Use It on Your Own Genome
A practical guide to the Genome Analysis Toolkit: what each tool does, which commands to run on a 30x whole genome, and where the pipeline fails.
17 posts
A practical guide to the Genome Analysis Toolkit: what each tool does, which commands to run on a 30x whole genome, and where the pipeline fails.
ReadPosRankSum is a Mann-Whitney U test comparing where reference and alternate alleles sit within the reads supporting them. This page explains the annotation, the thresholds we use on a personal genome, and why the value is missing at a third of your variants.
A practical guide to pulling SNPs, indels, samples, regions, and quality-filtered subsets out of your own whole-genome VCF with GATK SelectVariants, including the normalization step that prevents silent variant loss.
A step-by-step guide to producing a filtered germline VCF from your own 30x whole-genome FASTQs or CRAM, using GATK HaplotypeCaller, VQSR or hard filters, and concrete quality checks.
A step-by-step guide to going from FASTQ to a filtered, normalized VCF for a single human genome using bwa-mem2 and GATK4, with the flags, resource files, and sanity checks that matter.
A step-by-step guide to running gatk VariantFiltration on a personal VCF, with real thresholds, JEXL expression gotchas, genotype-level filters, and checks that tell you whether the filtering worked.
A step-by-step guide to applying and tuning GATK hard filters on a single-sample whole-genome VCF, including how to check your thresholds against a truth set instead of trusting the defaults.
A step-by-step guide to going from FASTQ or CRAM to a filtered, annotated germline VCF on one human genome using BWA-MEM2, GATK4 HaplotypeCaller, and hard filters, with the failure modes that matter at n=1.
A working pipeline for human exome data with BWA-MEM, GATK HaplotypeCaller and VEP, with the flags that matter and the failure modes.
A working pipeline that turns 30x whole-genome FASTQs into a filtered and benchmarked VCF, plus what changes when the variants are somatic.
A working pipeline from FASTQ to annotated VCF for one human genome, with DeepVariant and GATK calling and benchmarking against GIAB truth sets.
A working GATK pipeline for calling SNPs and indels from RNA-seq reads, with filters suited to transcriptomic data and a check against genome calls.
A working end-to-end path from raw FASTQ to an annotated, QC'd variant set using GATK, DeepVariant, and VEP, with the flags, file sizes, runtimes, and failure modes that matter.
A step-by-step pipeline from raw FASTQ to a filtered, benchmarked VCF for a single human genome, with the tool choices, parameters, and quality checks we would use.
A working pipeline for calling somatic SNVs and indels with GATK Mutect2, from BAM to annotated VCF. Covers results at low variant allele fraction.
A practical comparison of DeepVariant, GATK HaplotypeCaller, bcftools mpileup, Strelka2, and Clair3 for calling SNPs from your own whole-genome data, with commands, benchmarks, and the failure modes that matter.
A working end-to-end guide to going from FASTQ to a filtered, benchmarked VCF on a personal 30x human genome, with the specific tools, flags, reference files, and failure modes we would use.