Hello,
I have an array of count (quantification results with DESeq of RNA seq data). This table contains 40 columns (samples) and 33000 rows (a mixture between mRNAs and lncRNAs) I want to calculate the correlation between all lncRNAs and mRNAs, but I have no idea to start. If anyone can help me how to write the script I will be very grateful.
When you say "correlation between all lncRNA and mRNAs" do you intend to take all the measurements in the 40 samples for all lncRNA and all the mRNA measurements and see how correlated they are?
If so, a traditional correlation analysis cannot be performed because you will need the number of lncRNA measurements to equal the number of mRNA measurements.
If you want to look at pairwise correlations between all combinations of mRNA and lncRNAs, if you are using R you can either use the traditional cor function, or go for an optimised derivative like those in the WGCNA R package.
•
written
by
ankurchakravarthy •
thank you for answer
yes I am working on R but with the function corr it calculate correlation between the columns (sample1, sample 2, ......, smples 40) and not rows.
when I separated the tables (one for mRNAs and lncRNAs in another) I used the horn function gives an error message (the two tables must contain the same size)
Try transposing the matrix to get correlations between rows.
Mat <- t(Mat)
Log in to answer this question.
