Skip to content


CAVIAR (CAusal Variants Identification in Associated Regions), a statistical framework that quantifies the probability of each variant to be causal while allowing with arbitrary number of causal variants.

CAVIAR is available as a module on Apocrita.


To run the default installed version of CAVIAR, simply load the caviar module:

$ module load caviar

-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 (default 0.95)
-g GAMMA, --gamma       set $gamma$ the prior of a SNP being causal (default 0.01)
-c causal           set the maximum number of causal SNPs
-f 1                to out the probaility of different number of causal SNP

Long options currently do not work

The help documentation suggests that CAVIAR supports long options (i.e. --ld_file and --z_file however, using these options causes the application to crash with a "CAVIAR: invalid option -- '-' Strange error. The developers are aware of this, tracked in open issue #2).

CAVIAR is a single threaded application

CAVIAR does not provide support for multi-threading so please only request 1 core in your submission scripts. Any additional cores requested will be unused and wasted.

For usage documentation, run CAVIAR -h.

Example jobs

Serial job

Here is an example job running CAVIAR on 1 core and 1GB total memory:

#$ -cwd
#$ -j y
#$ -pe smp 1
#$ -l h_rt=1:0:0
#$ -l h_vmem=1G

module load caviar

CAVIAR -l pair_wise_correlation.ld \
       -z z_score_file.z

Here is an example job running eCAVIAR on 1 core and 1GB total memory:

#$ -cwd
#$ -j y
#$ -pe smp 1
#$ -l h_rt=1:0:0
#$ -l h_vmem=1G

module load caviar

eCAVIAR -l pair_wise_correlation1.ld \
        -l pair_wise_correlation2.ld \
        -z z_score_file1.z \
        -z z_score_file2.z \
        -o output_file