Hi everyone, I currently have a count matrix of Rna-seq data. To be precise I have 200 Tumor samples and 300 Healthy . Of those 200 Tumor samples I have 59 unique ones and the other are just the same sample with an additional string ("t0" in case is the Rna-seq of patient at time 0 , "t1" in case of treatment at time 1 and same for the other times until time 10 ). The problem is that not all samples have 10 times, meaning that Sample1 could have time 0 to time 3 , Sample 2 could have time 0 to time 8 and so on. What I want to study is how rna-seq of this tumor changes over treatment time . So I ask , just in words, how would you manipulate the data in order to reach this goal? What I need is just an idea on how would you manipulate the data and which design because so far I only did Differential Expression and Differential Co-Expression between Conditions, never studied time treatment development.
Little example of data :
Rows = Gene1 Gene2 Gene3 .... GeneN
Cols = Sample1t0 Sample2t0 .... Sample1t1 Sample2t1..... Sample1t2 Sample2t2.... ....Sample1t10 Sample2t10
Section 4.8 of the edgeR vignette features a detailed explanation of how to conduct time-course analysis of RNA-seq samples.
Hi thanks for the link, I have seen it and there is something I don't understand. For example in 4.8.1 he pass from 147 to 30 (unique) but when I check the sample information I see that for each 2 hours he has exactly 5 samples ("Embryos0002" has 5 samples, "Embryos0204" has 5 samples ...) , so the question is , are those 5 samples the same but just in different hours? Because if it is like that , it means that I must use samples that have the same time sets, so for instance If I want to study the time from 0 to 3 I must use the same samples for each time from 0 to 3 right? And they must be the same. I give you a better example of my interpretation:
This is right to do : Alex-t1 , Maria-t1, John-t1 , Alex-t2, Maria-t2, John-t2, Alex-t3, Maria-t3,John-t3 --> t1,t2,t3
This is not right to do : Alex-t1 , Maria-t1, John-t1, Alex-t2, Marco-t2, Luigi-t2, Alex-t3, Sasha-t3,Luisa-t3 --> Not good because different samples ?
If is like that then I understood and I close this topic otherwise how it is?