wes_skat (Deprecated!!)

==> Efficient solution to run SKAT, SKAT-O and Burden test are available on SAIGE-GENE+ <==

Nextflow workflow to run the SNP-set Kernel Association Test (SKAT) on whole-exome data (WES)

SKAT [1] is an emerging statistical method to test for association between variants/genes and continuous/binary phenotypes. The original SKAT R-package is single-thread based, meaning that the runtime can be long when analysing large dataset (thousands of genes/features) and multiple conditions (variant types and phenotypes).

Here, it is offered a work-around solution based on Nextflow (https://www.nextflow.io), following the split-apply-combine paradigm. The version presented here has been used for analysing large-scale exome-sequencing case-control cohorts, using the SKAT-O method.

Install/Dependences

Nextflow installation: curl -s https://get.nextflow.io | bash

See more info about Nexflow here.

R-dependences:

R >= 3.5.2, R-package ‘SKAT’ v2.2.2, R-package ‘optparse’ v1.6.6

Useage

Run SKAT to generate variant stats as showed below.

Command:

./nextflow run main.nf --snp_file path/to/snp_id/file \
                       --plink_input_file path/to/plink/files \
                       --n_sets 10 \
                       --run_variants

Parameters:

snp_file: Path to SNP id file. Expected a TSV file with two-columns, no-header. First column refers to genes/features, second to SNP IDs.

plink_input_file: Path to PLINK-formatted file(s) basename (without the extensions bim, bam and fam).

n_sets: Number of sets/chuncks to be generated.

run_variants: If set, run SKAT at variant level. Otherwise, run it at gene level (default).

Terminal output:

N E X T F L O W  ~  version 21.04.3
Launching `main.nf` [awesome_archimedes] - revision: 1586fead8e
executor >  local (11)
[db/c534db] process > generateSNPsets (1) [100%] 1 of 1 ✔
[b5/034e96] process > runSKAT (6)         [100%] 10 of 10 ✔
Results from SKAT test wrote to .../6d/merged_skat_results.csv

Results

Example output SKAT gene stats here.

Example output SKAT variant stats here.

References

[1] Zhao Z, Bi W, Zhou W, VandeHaar P, Fritsche LG, Lee S. UK Biobank Whole-Exome Sequence Binary Phenome Analysis with Robust Region-Based Rare-Variant Test. Am J Hum Genet. 2020 Jan 2. doi: 10.1016/j.ajhg.2019.11.012. PMID: 31866045.