Author: Sherine Awad
A snakemake pipeline to run various types of rare variants association analysis tests, including burden test, adaptive burden test, and more. Instead of testing whether a single variant is associated with a specific phenotype, we can study the cummulative effect of a group of variants in a gene or region on a specific phenotype. Rare variants association analysis has lots of categories, we will summarize here a few categories:
Burden test is prefered when we have variants in the same direction and the same effect, i.e. variants are risk increasing or risk decreasing. Many methods fall under this category including CMC, CMAT, etc.
Adaptive burden test is more robust than Burden test as it allows for adjusting weights. KBAC is famous approach that fall under the Adaptive burden test category.
Variance Components tests are powerful if there exist both trait-increasing and trait-decreasing variants or a small fraction of causal variants.
These types of tests are less powerful when the variants are in the same direction. A famous approach under this category is SKAT.
When we don't have any prior information about the variants we have, we can use combined test. These approaches are robust regardless of the percentage of causal variants and the existence of both trait-increasing and trait-decreasing variants. A famous approach under this category is SKATO.
You will need to edit the config file as below:
| Config Variable | Description |
|---|---|
| COHORT | Name of your Cohort |
| GENEFILE | Annotation file path |
| UPPER_FREQ | Maximum Frequency of the variant |
| OUTPUT | Name of your output |
The pipeline requires snakemake and conda installed.
You will need to have a .vcf and .ped file in your directory named as the COHORT field in the config file. You can edit this name in the configfile. You will need to edit the path of the gene file in the config file, as well as the output name and minimum frequency threshold required.
Then you use:
snakemake -jnumber_of_core
For example for 5 cores use:
snakemake -j5
For a dry run use:
snakemake -j1 -n
To print the commands in a dry run use:
snakemake -j1 -n -p
For the sake reproducibility, use conda to pull same versions of tools. Snakemake and conda have to be installed in your system:
snakemake --cores --use-conda
Add more tools as we go