GSCAN dbGaP

Framingham

ID Mapping

The following snppit describes how the PLINK genotype file IDs (which use "sampid") were mapped to the dbGaP_Subject_ID contained in the available phenotype files. The snippet only contains one of Joyce Ling's phenotype files for illustration only.

This script can be found in /work/KellerLab/GSCAN/dbGaP/Framingham/ID_mapping.r

### ID mapping file from dbGaP_Subject_ID to SAMPID 
ID.map <- read.table(gzfile("/work/KellerLab/GSCAN/dbGaP/Framingham/PhenoGenotypeFiles/RootStudyConsentSet_phs000007.Framingham.v28.p10.c1.HMB-IRB-MDS/PhenotypeFiles/phs000007.v28.pht001415.v16.p10.Framingham_Sample.MULTI.txt.gz"), header=T, sep="\t", stringsAsFactors=F)

### Genotype files
genotype.IDs <-  read.table("/work/KellerLab/GSCAN/dbGaP/Framingham/PhenoGenotypeFiles/ChildStudyConsentSet_phs000342.Framingham.v16.p10.c1.HMB-IRB-MDS/GenotypeFiles/phg000006.v9.FHS_SHARe_Affy500K.genotype-calls-matrixfmt.c1/subject_level_PLINK_sets/FHS_SHARe_Affy500K_subjects_c1.fam", header=F)
names(genotype.IDs) <- c("famid", "SAMPID", "patid", "matid", "sex", "phenotype")

length(which(genotype.IDs$SAMPID %in% ID.map$SAMPID))
## 6954 

x <- merge(genotype.IDs, ID.map, by="SAMPID", all.x=T)
x <- x[,c(1:5,7)]

### One of Joyce's phenotype ID files (eventually all phenotypes for all available participants were merged in, but this is a good example.)
phenotypes <- read.table("OffC_Exam_1.txt", header=T, sep="\t")

xx <- merge(x, phenotypes, by="dbGaP_Subject_ID", all=TRUE)

xx <- xx[,c(2:ncol(xx),1)] ## Reorder so the dbGaP_Subject_ID is the last column 

write.table(xx, file="framingham_GSCAN_phenotypesCovariates.ped", quote=FALSE, sep="\t", row.names=F)