In SVA, the percent of biological variable of interest didnot change after adding SV1.
Entering edit mode
3.1 years ago
dhwani2410 • 0

SVM plots "Pre SVM" "Post SVM"

I used SVM to find hidden confounding variable.The percentage of residuals were reduced after addition of new SV1 (can be seen in Post SVM image). However the percent of contribution of biological variable of interest("Pheno") still didnt go up.

Ideally the SV1 should increase the percentage of biological variable of interest. In such case should we add the SV1 as covariate while Differential expression analysis?

Also if i were to put this confounding variable in Combat function to get final matrix. what is the correct way to do it?

pheno_file=structure(list(Pheno = c("Vehicle", "Vehicle", "Vehicle", "Vehicle", 
"D0", "D0", "D0", "D0", "D2", "D2", "D2", "D2", "D2", "D4", "D4", 
"D4", "D4"), date = c("17", "17", "31", "31", "17", "31", "17", 
"31", "17", "17", "17", "31", "31", "17", "17", "31", "31"), 
    sv1_cat = c(1, 4, 1, 3, 3, 4, 3, 1, 2, 3, 4, 2, 2, 1, 4, 
    2, 1)), class = "data.frame", row.names = c("VN-170_(MTA-1_0).CEL", 
"VN-171_(MTA-1_0).CEL", "VN-172_(MTA-1_0).CEL", "VN-173_(MTA-1_0).CEL", 
"VN-174_(MTA-1_0).CEL", "VN-175_(MTA-1_0).CEL", "VN-176_(MTA-1_0).CEL", 
"VN-177_(MTA-1_0).CEL", "VN-178_(MTA-1_0).CEL", "VN-179_(MTA-1_0).CEL", 
"VN-180_(MTA-1_0).CEL", "VN-181_(MTA-1_0).CEL", "VN-182_(MTA-1_0).CEL", 
"VN-183_(MTA-1_0).CEL", "VN-184_(MTA-1_0).CEL", "VN-185_(MTA-1_0).CEL", 

modcombat = model.matrix(~1, data=pheno_file)
combat_edata = ComBat(dat=exprs(rma_Data), batch=pheno_file$Pheno, mod=modcombat,
                            par.prior=TRUE, prior.plots=FALSE)

Does it anywhere use the SV1 in giving the output?. Can someone please clarify this?

sva R combat • 777 views
Entering edit mode

Login before adding your answer.

Traffic: 2180 users visited in the last hour
Help About
Access RSS

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6