which is better for rna-seq analysis? R or python?

Question

[amirnavidinia2014]

which is better for rna-seq analysis? R or python? which is better for rna-seq analysis? R or python?

Answer 1

[ponganta]

I'd say it depends on your coding experience, although R is the go-to for any professional pretty useful (thanks ATpoint) for analysis at the moment.

PYTHON The syntax of python is much clearer for the beginner IMO. Most packages rely heavily on object oriented programming (OOP), so its much more intuitive to some folks. Also many R-packages have been ported to python (ggplot2 for example). However, most packages doesn't mean all (far from it!).

R is much more versatile than many people assume - simply because its primary purpose is data analysis. Its less fiddely than numpy, pandas etc. Once you learn how to use R, you can do basically everything - there will be a package for it.

I'm currently doing the following: Mining huge amounts of raw data? Do it in python. Data munging? Do it in R. One way to circumvent R's memory problems would be to simply use a combination of the holy trinity Bash -> Python -> R.

Edit: With "it depends on your coding experience" I didn't want to imply that only beginners are using Python. It was meant along the lines of "It depends on what workflows you were using previously". Sorry for my strong wording.

[ATpoint]

although R is the go-to for any professional analysis at the moment.

That is quite an overstatement. It implies that analysis outside of R is not professional and this is plainly wrong. R might be your preference, but please you should at least provide references or examples if you make strong statements like this. Look e.g. at the scRNA-seq field, there is indeed lots of R software, but you if you are into Py, then use e.g. something like scanpy to perform a full and perfectly valid analysis without even touching R.

[ponganta]

Thanks for the reminder - I agree that my wording is incorrect. R is not the only way to perform professional analyses.

Answer 2

[Thomas]

All the best packages for RNA-Seq analysis are written for R rather than python - e.g. Sleuth, DeSeq2 etc.

R is primarily designed for data analysis. Python isn't. And I am of the opinion that it is always better to work with your programming language, rather than use that language for something it wasn't originally designed to do. This is just my opinion, and I know others disagree however.

NB: I am talking about bulk RNA-Seq analysis here. I have no idea what the state of play is for single cell RNA-Seq analysis.

Answer 3

[huynguyen96.dnu]

I think this question is kind of can-not-answer one. It depends greatly on your interest and expertise in either R or Python.