Skip to content

Latest commit

 

History

History
44 lines (39 loc) · 2.15 KB

README.md

File metadata and controls

44 lines (39 loc) · 2.15 KB
Raw

finemapping

A Nextflow framework for finemap

Installation:

Download this pipeline locally or on the medcluster.
Install finemap_v1.4_x86_64.tgz within the bin folder so finemap is located in bin/finemap_v1.4_x86_64/finemap_v1.4_x86_64.

Usage:

Pipeline needs 6 input files:
Reference: in bim, bed, fam (3 files with same basename)
Locus-file: csv-file setting the boundaries of the finemap plots; columns: chunk,NSNP,chr,st,sp,PPA_3

  • chunk: iterate your loci like 1,2,3...
  • NSNP: currently not in use, put in NA
  • chr: chromosome, like 12, not chr12, and 23 instead of X
  • st: coordinate where plotting starts
  • sp: coordinate where plotting ends
  • PPA_3: currently not in use, put in 1

SNP-List: file with 1 snp per row, all other SNPs are excluded drom plotting and finemapping
SUMSTAT-FILE: file containing following columns:
CHR BP SNP A1 A2 P OR BETA SE N CHISQ Z SOURCE FRQ_A_A1 FRQ_U_A1 INFO
only CHR, BP, SNP, A1, A2, P, BETA, SE, FRQ_U_A1 are relevant, the other columns can be filled with NA. Call pipeline with:

nextflow run main.nf -profile standard --locus /home/user/finepipe/example/locusfile.sample --snps /home/user/finepipe/example/snplist.sample --reference /home/user/finepipe/example/GerNorItaSpa.chr3 --sumstats /home/user/finepipe/example/sumstats.sample --nsum 15743 --nsignal 1 --method sss -resume

Mandatory:

--locus "path/to/locus.file"
--snps "path/to/snp.list"
--reference "path/to/bimbedfam" (no file extension)
--sumstats "path/to/sumstat.file"
--nsum N of studysize
--method "sss" or "cond"
--nsignal N of max signals

If you run it locally, download the locuszoom database and set it with:
--locuszoomdb "/path/to/locuszoom/data/database/locuszoom_hg38.db"
and set profile to local:
-profile "local" or "standard"

Optional:

--dprime sets ld method from the default r² to dprime
--output "/path/to/output" if not set output is baseDir/Results
-resume Continue a run with cached processes