Skip to content

Home

Jump to bottom
Hae Kyung Im edited this page Jul 29, 2020 · 12 revisions

Help!

We desperately need someone to document this stuff!

In the meantime, use this tutorial in this repo's wiki and another one in the MetaXcan wiki. Additional options used by the GTEx GWAS subgroup here can be helpful.

You can find a wide array of assorted command line tools to perform genomic analysis on different analysis. Most tools are concerned with integration of GWAS summary statistics and eQTLs.

Imputing GWAS summary statistics

We make available here a complete tutorial, using a completely functional and freely available 1000 Genomes-based genotype panel (here).

Our application to GTEx data is detailed in the following posts:

  1. assemble a genotype reference
  2. Impute summary statistics missing variants Unfortunately, the underlying data can't be shared freely. These articles are presented as reference for people interested in using their own data.

Running Coloc

You can compute coloc's results on all genes for a GTEX-like eQTL study and a gwas. The tool herein assumes GWAS in the format generated by the standard settings from the imputation tools in the previous section, and then Run coloc as in this entry.

Dependencies

GWAS summary imputation tools were developed on python 3.5. Later versions of python work but are not tested.

Here are the specific versions of python packages used when the software was developed:

h5py==2.8.0
numpy==1.15.4
pandas==0.23.4
patsy==0.5.1
pyarrow==0.9.0
pyliftover==0.3
rpy2==2.9.4
scipy==1.1.0
statsmodels==0.9.0

pyarrow 0.10.0 should also work, but users have reported issues with the latest pyarrow due to API changes.

Clone this wiki locally