Editing GSCAN dbGaP (section)

==eMERGE==

===Phenotypes===
<syntaxhighlight lang="rsplus">

options(stringsAsFactors=F)


### eMERGE is broken into different consent classes. We can conduct analyses on hmb, hmb-gso-nic, and 


emerge.hmb <- read.table("/work/KellerLab/dbGaP/eMERGE-MergedSet/PhenoGenotypeFiles/RootStudyConsentSet_phs000360.eMERGE_MergedSet.v3.p1.c1.HMB/PhenotypeFiles/phs000360.v3.pht003255.v2.p1.c1.MergedSet_Subject_Phenotypes.HMB.txt.gz", header=TRUE, sep="\t", stringsAsFactors=F)
emerge.hmb.genos <- read.table("/work/KellerLab/dbGaP/eMERGE-MergedSet/PhenoGenotypeFiles/RootStudyConsentSet_phs000360.eMERGE_MergedSet.v3.p1.c1.HMB/GenotypeFiles/matrix/c1.HMB/eMerge_660_11212012_c1.fam", header=FALSE, sep="\t", stringsAsFactors=F)


emerge.hmb.gso.nic <- read.table("/work/KellerLab/dbGaP/eMERGE-MergedSet/PhenoGenotypeFiles/RootStudyConsentSet_phs000360.eMERGE_MergedSet.v3.p1.c3.HM-B-GSO-NIC/PhenotypeFiles/phs000360.v3.pht003255.v2.p1.c3.MergedSet_Subject_Phenotypes.HM-B-GSO-NIC.txt.gz", header=TRUE, sep="\t", stringsAsFactors=F)
emerge.hmb.gso.nic.genos <- read.table("/work/KellerLab/dbGaP/eMERGE-MergedSet/PhenoGenotypeFiles/RootStudyConsentSet_phs000360.eMERGE_MergedSet.v3.p1.c3.HM-B-GSO-NIC/GenotypeFiles/matrix/c3.HM-B-GSO-NIC/eMerge_660_11212012_c3.fam", header=FALSE, sep="\t", stringsAsFactors=F)


emerge.hmb.gso <- read.table("/work/KellerLab/dbGaP/eMERGE-MergedSet/PhenoGenotypeFiles/RootStudyConsentSet_phs000360.eMERGE_MergedSet.v3.p1.c4.HMB-GSO/PhenotypeFiles/phs000360.v3.pht003255.v2.p1.c4.MergedSet_Subject_Phenotypes.HMB-GSO.txt.gz", header=TRUE, sep="\t", stringsAsFactors=F)
emerge.hmb.gso.genos <- read.table("/work/KellerLab/dbGaP/eMERGE-MergedSet/PhenoGenotypeFiles/RootStudyConsentSet_phs000360.eMERGE_MergedSet.v3.p1.c4.HMB-GSO/GenotypeFiles/matrix/c4.HMB-GSO/eMerge_660_11212012_c4.fam", header=FALSE, sep="\t", stringsAsFactors=F)


### Merge all files above according to SUBJID, which is used in the
### genotype files.

emerge <- merge(emerge.hmb, emerge.hmb.gso, all=T)
emerge <- merge(d, emerge.hmb.gso.nic, all=T)

### SMOKING INITIATION
###
### The eMERGE variable name is SMOKING_STATUS
###      C65108 = never smoker
###      C67147 = current smoker
###      C67148 = past smoker
###      C67151 = Unknown if ever smoked
###
### Descriptives:
###
### table(emerge$SMOKING_STATUS)
###
### C65108 C67147 C67148 C67151 
###   2217   1736   3457   9635 

si <- emerge$SMOKING_STATUS
si[si == "C67147" | si == "C67148"] <- 2
si[si == "C65108"] <- 1
si[si != 1 & si != 2] <- NA

### SMOKING Cessation
###
### Current == 2 & Former == 1 in GSCAN. This is already the case for these data.

sc <- emerge$SMOKING_STATUS
sc[sc == "C67147"] <- 2
sc[sc == "C67148"] <- 1
sc[sc != 1 & sc != 2] <- NA


### eMERGE age variable is tricky because there is no obvious age at
### assessment. We will use their "DECADE_BIRTH" as a terrible
### approximation.
### 1=1900-1919; 2=1920-1929, 3=1930-1939; 4=1940-1949; 5=1950-1959; 6=Unknown
###
### Descriptives:
###
### table(emerge$DECADE_BIRTH)
###
###   .    1    2    3    4    5    6    7    8    9   99 
###   6  612 2667 3533 4439 3127 1291  761  490   10  109 
birthyear <- emerge$DECADE_BIRTH
birthyear[birthyear == "99"] <- NA
birthyear[birthyear == "."] <- NA


### SEX
sex <- emerge$SEX
sex[sex == "C46109"] <- 1
sex[sex == "C46110"] <- 2


### Scott decided not to correct for additional case-control variables
### given what appears to be a highly complex sample and uncertainty
### about the best course of action to account for disease status in
### conducting smoking analyses.

phenotypes <- data.frame(fid = emerge$SUBJID,
                         iid = emerge$SUBJID,
                         patid = "x",
                         matid = "x",
                         sex = sex,
                         si = si,
                         sc = sc)

phenotypes[is.na(phenotypes)] <- "x"

write.table(phenotypes,
            "/work/KellerLab/vrieze/GSCAN/GWAS/summary_stats_generated_internally/eMERGE/GSCAN_eMERGE_phenotypes.ped",
            row.names=F,
            quote = F,
            sep="\t")


covariates  <- data.frame(fid = emerge$SUBJID,
                          iid = emerge$SUBJID,
                          patid = "x",
                          matid = "x",
                          sex = sex,
                          birthyear = birthyear)                         

covariates[is.na(covariates)] <- "x"

write.table(covariates,
            "/work/KellerLab/vrieze/GSCAN/GWAS/summary_stats_generated_internally/eMERGE/GSCAN_eMERGE_covariates.ped",
            row.names=F,
            quote = F,
            sep="\t")
</syntaxhighlight>