Survival analysis: data clinical and pathways

Question

[fp89]

Hello, I need to perform survival analysis to find significant associations of specific pathway activities to patient survival. I'm trying to perform the analysis using this tutorial https://github.com/mforde84/RNAseq-Survival-Analysis-TCGA-KIRC/blob/master/survival_rnaseq_analysis.R. I have two files: clinical data and uni_vals. There is something wrong I think in the part of survival analysis. The message is the following:"Error in time[[i]] <- sort(unique(y[who, 1])) : attempt to select less than one element in integerOneIndex"

1)clinical data

structure(list(Tumor_Sample_Barcode = structure(c(126L, 128L, 133L), .Label = c("TCGA-A7-A0CE", "TCGA-A7-A0CH", 
"TCGA-A7-A0DC"), class = "factor"), 
       classification_of_tumor = structure(c(1L, 1L, 1L), .Label   = "not reported", class = "factor"), 
last_known_disease_status = structure(c(1L, 1L,1L
), .Label = "not reported", class = "factor"), updated_datetime = structure(c(1L, 
1L, 1L), .Label = "2018-01-19T13:39:21.801433-06:00", class = "factor"), 
primary_diagnosis = structure(c(6L, 6L, 6L), .Label = c("C50.2", 
"C50.3", "C50.4", "C50.5", "C50.8", "C50.9", "C50.919"), class = "factor")), row.names = c(NA, 5L), class = c("data.table", "data.frame"))

2)uni_vals

structure(list(`TCGA-A7-A0CE` = c(0.800945270510658, 0.99887793401069, 
0.667341683672123, 0.999999999314536, 0.999999999314536), `TCGA-A7-A0CE.1` = c(0.778700980142054, 
0.998594728888895, 0.762898025094707, 0.999999999620033, 0.999999999620033), `TCGA-A7-A0CH` = c(0.608118239725987, 0.992929539569249, 0.706256002082062, 0.999999998256691, 0.999999998256691), `TCGA-A7-A0CH.1` = c(0.899309224249365, 0.999869380713655, 0.797778413011216, 0.9999999997944, 0.9999999997944), `TCGA-A7-A0DC` = c(0.535342987646728, 0.993464915142776, 0.409818699577936, 0.999999996633627, 0.999999996633627)), row.names = c("Lipid degradation", "Lipid metabolism", "Chemotaxis", "Transcription regulation", "Transcription"), class = "data.frame")

3) my code

get the index of the normal/control samples

n_index <- which(substr(colnames(uni_vals),14,14) == '1')
t_index <- which(substr(colnames(uni_vals),14,14) == '0')

input clinical information

all_clin < -data.frame(cbind(clinical[,7],clinical[,10],clinical[,22]))
colnames(all_clin) <- c("new_tumor_days", "death_days", "followUp_days")
rownames(all_clin) <- clinical$Tumor_Sample_Barcode

time to tumor

all_clin$new_time <- c()
for (i in 1:length(as.numeric(as.character(all_clin$new_tumor_days)))){all_clin$new_time[i] <- ifelseis.na(as.numeric(as.character(all_clin$new_tumor_days))[i]), 
                             as.numeric(as.character(all_clin$followUp_days))[i],
                             as.numeric(as.character(all_clin$new_tumor_days))[i])

}

time to death

all_clin$new_death <- c()
for (i in 1:length(as.numeric(as.character(all_clin$death_days)))){
  all_clin$new_death[i] <-ifelseis.na(as.numeric(as.character(all_clin$death_days))[i]),
                               as.numeric(as.character(all_clin$followUp_days))[i],
                              as.numeric(as.character(all_clin$death_days))[i])

}

death censor event

all_clin$death_event <- ifelse(clinical$vital == "alive", 0, 1)

filtering

colnames(uni_vals) <- gsub("\\.","-",substr(colnames(uni_vals),1,12))

match all_clin and uni_vals

ind_tum <- which(unique(colnames(uni_vals)) %in% rownames(all_clin))
ind_clin <- which(rownames(all_clin) %in% colnames(uni_vals))

function of interest

ind_func <- which(rownames(uni_vals) == "Lipid degradation")

create event vector for uni data

event_uni <- t(apply(uni_vals, 1, function(x) ifelse(abs(x) <0.01,1,0)))

survival analysis

s <- survfit(Surv(as.numeric(as.character(all_clin$new_death))[ind_clin], 
              all_clin$death_event[ind_clin]) ~ event_uni[ind_func, ind_tum])
s1 <- tryCatch( survdiff(Surv(as.numeric(as.character(all_clin$new_death))[ind_clin], 
            all_clin$death_event[ind_clin]) ~ event_uni[ind_func, ind_tum]), 
  error = function(e) 
return(NA)

)

Answer 1

[zx8754]

The error is explained at SO:

• Error in lis[[i]] : attempt to select less than one element

You are accessing 0th index of a list. In R indexing starts at 1. Try to find out where this is happening in your code or on that GitHub code.