======
a statistical framework that quantifies the probability of each variant to be causal while allowing with arbitrary number of causal variants.
original publications can be found here for eCAVIAR and here for CAVIAR.
Developed at the ZarLab at UCLA
More information on CAVIAR and eCAVIAR can be found on the CAVIAR website
to install this repository --
git clone https://github.com/fhormoz/caviar.git
./CAVIAR [options]
Options:
-h, --help show this help message and exit
-o OUTFILE, --out=OUTFILE specify the output file
-l LDFILE, --ld_file=LDFILE the ld input file
-z ZFILE, --z_file=ZFILE the z-score and rsID files
-r RHO, --rho-prob=RHO set $pho$ probability
-c causal set the maximum number of causal SNPs
-f 1 to out the probaility of different number of causal SNP
Usage: ./eCAVIAR [options]
Options:
-h, --help show this help message and exit
-o OUTFILE, --out=OUTFILE specify the output file
-l LDFILE, --ld_file=LDFILE the GWAS ld input file
-l LDFILE, --ld_file=LDFILE the eQTL ld input file
-z ZFILE, --z_file=ZFILE the GWAS z-score and rsID files
-z ZFILE, --z_file=ZFILE the eQTL z-score and rsID files
-r RHO, --rho-prob=RHO set $pho$ probability
-c causal set the maximum number of causal SNPs
-f 1 to out the probaility of different number of causal SNP
OUTFILE_1_set - causal SNP in GWAS
OUTFILE_2_set - causal SNP in eQTL
OUTFILE_1_post - Causal posterior probability for each SNP in GWAS
OUTFILE_2_post - Causal posterior probability for each SNP in eQTL
OUTFILE_col - The Colocalization posterior probability (CLPP) for each SNP.
CAVIAR is written in C++ and must be compiled before running. If you are encountering errors in running CAVIAR or eCAVIAR try these steps:
- check if you have the GNU scientific library installed
* for macOS this can be done using the homebrew package manager-
brew install gsl
- Next, in the caviar/CAVIAR C++ repository type
make clean
make
chmod +x eCAVIAR
may also be helpful CAVIAR should be able to run using these parameters
Other helpful hints - if running eCAVIAR make sure your LD files have the same SNPs, works best for low to medium LD other related code developed by UCSF students here
CAVIAR is offered under the GNU Affero GPL (https://www.gnu.org/licenses/why-affero-gpl.html).