Thank you for your answer. I think my lack of knowledge about the overall process was the reason to look for a step by step tutorial. However, because I already know how to perform QC and alignment, I think your answer is a short useful tutorial for me!! Don't I need to do differential expression analysis and just doing correlation analysis suffice? I saw a few papers with considerably complicated algorithms for finding miRNA-mRNA regulatory networks!!
I'm a wet lab Molecular Biologist who has experience with programming in Python and R, and I also know how to use bash. However, I'm a newbie in RNA seq data analysis, and I just read and coded through a few tutorials here and there.
Right now, I need to work on miRNA-mRNA negative correlation analysis, and I'm searching to find a code template or pipeline to use as my starting points instead of writing everything from scratch.
I also need a step by step tutorial, preferably in Jupyter notebook format, to learn the why behind different stages of differential analysis, enrichment analysis, and correlation analysis.
I appreciate your help
1 answer
Not sure you could find a step by step tutorial while you need only a simple correlation (Negative or positive) analysis. Hope this simple way can help you!
- Provide a read count table including your miRNA and mRNAs and import into Rstudio
- Calculate log2: log_counts <- log2(raw_counts + 1)
And finally, make a cor matrix using following function.
cordist <- function(dat) {cor_matrix <- cor(t(dat)) dist_matrix <- as.matrix(dist(dat, diag=TRUE, upper=TRUE)) dist_matrix <- log1p(dist_matrix) dist_matrix <- 1 - (dist_matrix / max(dist_matrix)) sign(cor_matrix) * ((abs(cor_matrix) + dist_matrix)/ 2)} sim_matrix <- cordist(log_counts)
Log in to answer this question.