tinglab / reCAT

20 stars 9 forks source link

How to get time series based known cycle labels? #7

Open yancychy opened 5 years ago

yancychy commented 5 years ago

Hi, I tired to use the training.data. cellid = colnames(training.data) cycles = sapply(cellid, function(x) { strsplit(x, "_")[[1]][1] }) timelab = sapply(cycles, function(x) { if(x=="G1"){ y=1 }else if(x=="S") { y=2 }else if(x=="G2M"){ y=3 }else{ y=0 }
y }) ordIndex <- get_ordIndex(training.data, threadnum=4) cor(ordIndex, timelab) The results is 0.62. Why the correlation is so small compared to the paper?

Thanks.