Joint genotype calling across multiple samples using GATK CombineGVCFs and GenotypeGVCFs. Essential for cohort studies, population genetics, and leveraging VQSR.
SkillScore95Maintained
Automatic
◆Clinical / Healthcare pack
Last verified 8/2/2026
What it changes
Passed 21 of 22 test cases with the skill, and 20 of 22 without it.
Without skill91%
With skill95%
Checks it passed that the model alone missed
The bcftools concat command includes -Oz compression output.
1 fixed20 passed either way1 still failing0 broke
Pass-rate lift
+5 pp
Tokens
+137%
Agent turns
0%
Response time
-36%
Model gemini-3.6-flashMeasured 8/2/2026v121 of 22 cases passed with skill
Tokens, turns and time describe resource use, not a dollar price.
See it in action
A test case the skill fixed. Same model, same prompt, with and without the skill.
Prompt
After running GenotypeGVCFs separately across chromosomes 1 through 22, X, and Y from GenomicsDB workspaces, we have 24 per-chromosome compressed VCF files (`chr1.vcf.gz` through `chrY.vcf.gz`). What command using bcftools should be used to concatenate these files into a single cohort file?
What was checked, without and with the skill
Checked
Without skill
With skill
The command uses bcftools concat to merge per-chromosome files.
Passed
Passed
The bcftools concat command includes -Oz compression output.
Missed
Passed
case-13 · gemini-3.6-flash · 8/2/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md332 linessingle file
Joint Calling
Call variants jointly across multiple samples for improved accuracy and consistent genotyping.
Why Joint Calling?
Improved sensitivity - Leverage information across samples
Consistent genotyping - Same sites called across all samples
VQSR eligible - Requires cohort for machine learning filtering
Population analysis - Allele frequencies across cohort
Workflow Overview
Sample BAMs
│
├── HaplotypeCaller (per-sample, -ERC GVCF)
│ └── sample1.g.vcf.gz, sample2.g.vcf.gz, ...
│
├── CombineGVCFs or GenomicsDBImport
│ └── Combine into cohort database
│
├── GenotypeGVCFs
│ └── Joint genotyping
│
└── VQSR or Hard Filtering
└── Final VCF