Clumping in GWASLab by calling PLINK2

GWASLab provides a wrapper for clumping using PLINK2. You can get clumping results without worrying about the file format.

Clumping

You need to install PLINK2 first and add to your environment.

clump(  plink2="plink2",
        vcf=None, 
        bfile=None,
        pfile=None，
        scaled=False, 
        out="clumping_plink2", 
        overwrite=False, 
        n_cores=2, 
        chrom=None, 
        clump_p1=5e-8, 
        clump_p2=5e-8, 
        clump_r2=0.2, 
        clump_kb=250,
        log=Log())

Option	DataType	Description	Default
plink2	string	path to plink2	'plink2'
vcf	string	path to reference VCF file (it will be converted to plink binary format). You need to specify either vcf ,bfile or bfile	-
bfile	string	path to PLINK bfile. You need to specify either vcf ,bfile or bfile	-
pfile	string	path to PLINK bfile. You need to specify either vcf bfile， or pfile	-
scaled	boolean	If tru, use MLOG10P instead of P	False
out	string	output file prefix	clumping_plink2
overwrite	boolean	if True, overwrite the existing bfile when vcf path is provided	False
n_cores	int	number of cores to use	2
memory	int	number of memory to use for plink (in MB)	None
clump_p1	float	clump_p1	5e-8
clump_p2	float	clump_p2	5e-8
clump_r2	float	clump_r2	0.2
clump_kb	float	clump_kb	250

Example

Example

clumps = mysumstats.clump(vcf= "/home/yunye/.gwaslab/EAS.ALL.split_norm_af.1kgp3v5.hg19.vcf.gz")

Brisbane plot

See Clumping in gwaslab

Plink2 script for clumping

# using P
plink2 \
    --bfile {}\
    --chr {} \
    --clump {} \
    --clump-field P \
    --clump-snp-field SNPID \
    --clump-p1 {} \
    --clump-p2 {} \
    --clump-r2 {} \
    --clump-kb {} \
    --threads {} \
    --out {}

or

# using MLOG10P
plink2 \
    --bfile {}\
    --chr {} \
    --clump {} \
    --clump-field P \
    --clump-snp-field SNPID \
    --clump-p1 {} \
    --clump-p2 {} \
    --clump-r2 {} \
    --clump-kb {} \
    --threads {} \
    --out {}