How do I subset a dataframe column based on conditions of another column?-CodePudding

For each column name with nM* prefix, I want to find each row with value <500 and store the corresponding Peptide row value as pep500. My code below returns an empty pep500 vector.

netmhc <- read.table("31917_NetMHC.xls", sep="\t", row.names=NULL, header=TRUE, skip=1)
# Find peptides with binding affinity <500nM
pep500 <- vector()
for(i in 1:ncol(netmhc)){
  if(grepl('nM', colnames(netmhc[i])==TRUE)){
    for(j in 1:nrow(netmhc)) {
      if(j>=500){
        skip
      }
      else{
        for(k in netmhc["Peptide"]){
          pep500[i] <- k
        }
      }
    }    
  }
}

Data structure

structure(list(Pos = c(0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 
0L, 0L, 0L, 0L, 0L), Peptide = c("VPVISPRIH", "SGRHNWEVG", "SGIHPGATT", 
"TELAKHQAI", "QHTHAIESN", "QVALEELRR", "SLALLELSV", "KFHATEENR", 
"STESEIPDY", "AALLQSKSN", "VEGLCYLTG", "RPGLWPAPN", "LNGLDQQMI", 
"EELQWIFPS", "GASPVPNPS"), ID = 1:15, nM = c(46125.8, 43001.7, 
30957.3, 11468.1, 45431.9, 46655.8, 166.5, 40224.7, 38301.1, 
42644, 44350.7, 29476.3, 20441.6, 9224.3, 28582.4), Rank = c(75, 
55, 23, 9.5, 70, 85, 1.5, 41, 35, 50, 60, 22, 14, 8, 21), Core = c("VPVISPRIH", 
"SGRHNWEVG", "SGIHPGATT", "TELAKHQAI", "QHTHAIESN", "QVALEELRR", 
"SLALLELSV", "KFHATEENR", "STESEIPDY", "AALLQSKSN", "VEGLCYLTG", 
"RPGLWPAPN", "LNGLDQQMI", "EELQWIFPS", "GASPVPNPS"), nM.1 = c(32273.4, 
24415.1, 37350.4, 1954.7, 39374.7, 39018.4, 8669.8, 42118.9, 
37011.7, 26852.3, 32381.1, 25391.5, 19678.8, 32284.2, 28078.5
), Rank.1 = c(60, 33, 80, 2.5, 90, 90, 8.5, 99, 80, 40, 60, 36, 
23, 60, 43), Core.1 = c("VPVISPRIH", "SGRHNWEVG", "SGIHPGATT", 
"TELAKHQAI", "QHTHAIESN", "QVALEELRR", "SLALLELSV", "KFHATEENR", 
"STESEIPDY", "AALLQSKSN", "VEGLCYLTG", "RPGLWPAPN", "LNGLDQQMI", 
"EELQWIFPS", "GASPVPNPS"), H_Avg_Ranks = c(66.667, 41.25, 35.728, 
3.958, 78.75, 87.429, 2.55, 57.986, 48.696, 44.444, 60, 27.31, 
17.405, 14.118, 28.219), N_binders = c(0L, 0L, 0L, 0L, 0L, 0L, 
1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L)), class = "data.frame", row.names = c(NA, 
-15L))

Expected output: pep500 = SLALLELSV

CodePudding user response：

We may use if_any in filter i.e. loop over the columns with names starts_with 'nM' prefix, check whether any of those columns have less than 500 in any columns in the looped

library(dplyr)
pep500 <-  df %>% 
    filter(if_any(starts_with("nM"), ~ .x < 500))

-output

pep500
  Pos   Peptide ID    nM Rank      Core   nM.1 Rank.1    Core.1 H_Avg_Ranks N_binders
1   0 SLALLELSV  7 166.5  1.5 SLALLELSV 8669.8    8.5 SLALLELSV        2.55         1