Same as computeFDRwithID but works with decoy_hit boolean vector.
For more details and references see package vignette
vignette("TargetDecoyFDR_Example", package = "prozor")
computeFDR(score, decoy_hit, larger_better = TRUE)
| score | score |
|---|---|
| decoy_hit | indicates if decoy hit |
| larger_better | is larger score the better one (default TRUE) |
list with decoy_hit (indicates if decoy), score the search engine score, FDR1 false discovery rate estimated using the method of Gygi, SimpleFDR - estimated using the method of Kaell.
data(fdrSample) fdr1 <- computeFDR(fdrSample$score, grepl("REV_",fdrSample$proteinID), larger_better = FALSE) head(as.data.frame(fdr1)) #> larger_better order decoy_hit score FPR SimpleFDR qValue_FPR #> 1 FALSE 287 FALSE 8.611747e-21 0 0 0 #> 2 FALSE 3581 FALSE 6.305185e-19 0 0 0 #> 3 FALSE 29584 FALSE 1.750698e-17 0 0 0 #> 4 FALSE 26559 FALSE 3.081044e-16 0 0 0 #> 5 FALSE 15524 FALSE 3.835269e-16 0 0 0 #> 6 FALSE 30840 FALSE 3.845918e-16 0 0 0 #> qValue_SimpleFDR #> 1 0 #> 2 0 #> 3 0 #> 4 0 #> 5 0 #> 6 0