QTL association testing

QTL association testing#

This mini-protocol selects and runs cis-, trans-, or interaction-QTL association testing with the TensorQTL module.

Miniprotocol Timing#

Timing: TBD

Overview#

QTL association testing identifies genetic variants associated with molecular traits such as gene expression. This mini-protocol calls TensorQTL.ipynb for three independent analyses: a cis scan of variants near each molecular trait, a trans scan of variants outside the local region, and an interaction scan testing whether a covariate modifies the genotype effect.

Run the genotype-, phenotype-, and covariate-preprocessing mini-protocols first. Their outputs provide chromosome-indexed PLINK genotypes, chromosome-indexed molecular phenotypes, and a sample-aligned covariate matrix. The association results can then be passed to association postprocessing, fine-mapping, or multi-omics integration.

Choose one route below; the three commands are alternatives rather than one mandatory chain.

Steps#

Choose one route; cis-, trans-, and interaction-QTL scans answer different questions and are not a mandatory chain.

Analysis goal	Command to run	Inputs
Test local genetic effects within each phenotype’s cis window	1	`output/genotype_by_chrom/protocol_example.genotype.merged.plink_qc.genotype_by_chrom_files.txt`; `output/phenotype/phenotype_by_chrom_for_cis/bulk_rnaseq.phenotype_by_chrom_files.txt`; `output/covariate/protocol_example.rnaseq.bed.protocol_example.covariates.protocol_example.genotype.merged.plink_qc.plink_qc.prune.pca.Marchenko_PC.gz`
Test distal effects across chromosomes	2	The same genotype, phenotype, and covariate files, plus `data/combined_AD_genes.csv` as the example phenotype-region list
Test whether a covariate modifies a cis-QTL effect	3	The same genotype, phenotype, and covariate files, plus the interaction variable `msex` in the covariate data

Run only the command matching the scientific question.

1. cis-QTL scan #

What it does: Tests each molecular trait against variants within its cis window, using --MAC 5 for the small chromosome 22 example.

sos run pipeline/TensorQTL.ipynb cis \
    --genotype-file output/genotype_by_chrom/protocol_example.genotype.merged.plink_qc.genotype_by_chrom_files.txt \
    --phenotype-file output/phenotype/phenotype_by_chrom_for_cis/bulk_rnaseq.phenotype_by_chrom_files.txt \
    --covariate-file output/covariate/protocol_example.rnaseq.bed.protocol_example.covariates.protocol_example.genotype.merged.plink_qc.plink_qc.prune.pca.Marchenko_PC.gz \
    --cwd output/tensorqtl_cis --name protocol_example --MAC 5 --numThreads 2

2. trans-QTL scan #

What it does: Tests the selected traits against variants on chromosome 22, restricting traits to the identifiers listed in data/combined_AD_genes.csv.

sos run pipeline/TensorQTL.ipynb trans \
    --genotype-file output/genotype_by_chrom/protocol_example.genotype.merged.plink_qc.genotype_by_chrom_files.txt \
    --phenotype-file output/phenotype/phenotype_by_chrom_for_cis/bulk_rnaseq.phenotype_by_chrom_files.txt \
    --covariate-file output/covariate/protocol_example.rnaseq.bed.protocol_example.covariates.protocol_example.genotype.merged.plink_qc.plink_qc.prune.pca.Marchenko_PC.gz \
    --cwd output/tensorqtl_trans --name protocol_example --MAC 5 --numThreads 2 \
    --trans-geno-chromosome 22 --region-list data/combined_AD_genes.csv --region-list-phenotype-column 4

3. interaction-QTL scan #

What it does: Runs the cis model with a genotype-by-msex interaction term and reports evidence that the genotype effect changes with this covariate.

sos run pipeline/TensorQTL.ipynb cis \
    --genotype-file output/genotype_by_chrom/protocol_example.genotype.merged.plink_qc.genotype_by_chrom_files.txt \
    --phenotype-file output/phenotype/phenotype_by_chrom_for_cis/bulk_rnaseq.phenotype_by_chrom_files.txt \
    --covariate-file output/covariate/protocol_example.rnaseq.bed.protocol_example.covariates.protocol_example.genotype.merged.plink_qc.plink_qc.prune.pca.Marchenko_PC.gz \
    --cwd output/tensorqtl_int --name protocol_example --MAC 5 --numThreads 2 \
    --interaction msex --maf-threshold 0.05 --no-permutation

Output Files#

Route	Relative path	Contents
cis-QTL	`output/tensorqtl_cis/protocol_example.<phenotype_chr>_<genotype_chr>.cis_qtl.pairs.tsv.gz`	Nominal statistics for tested cis variant-trait pairs
cis-QTL	`output/tensorqtl_cis/protocol_example.<phenotype_chr>_<genotype_chr>.cis_qtl_regional_significance.tsv.gz`	Region-level permutation and multiple-testing results
cis-QTL	`output/tensorqtl_cis/protocol_example.<phenotype_chr>.cis_qtl_pairs.<chromosome>.parquet`	Native TensorQTL nominal results retained for reuse
trans-QTL	`output/tensorqtl_trans/protocol_example.<phenotype_chr>_geno_chr22.trans_qtl.pairs.tsv.gz`	Trans-association statistics for chromosome 22
interaction-QTL	`output/tensorqtl_int/protocol_example.<phenotype_chr>_<genotype_chr>_msex.cis_qtl.pairs.tsv.gz`	Nominal genotype, interaction, and genotype-by-interaction effects

Tabix index files (.tbi) accompany the bgzipped result tables. Exact chromosome tokens follow the input file lists.

Anticipated Results#

The selected route produces association statistics for the chromosome 22 example. Cis analysis returns nominal variant-trait results plus region-level significance summaries; trans analysis returns the restricted cross-region tests; interaction analysis adds main and genotype-by-covariate effect estimates.

Proceed to association postprocessing when calibrated regional summaries are needed, or use the nominal and region-level tables as input to the appropriate fine-mapping or integration workflow.

Command interface#

Inspect all TensorQTL workflows, options, and defaults:

sos run pipeline/TensorQTL.ipynb -h