SEQanswers (
-   Bioinformatics (
-   -   Learning R or Python for genome comparisons (

RosanneHertzberger 05-14-2016 10:42 AM

Learning R or Python for genome comparisons
Hi, I'm new here, originally a wetlab microbiologist, now trying to get some bio-informatics skills. So here is my question:

I know that from one bacterial organism Lactobacillus crispatus, a few strains show a certain metabolism whereas other strain are not capable of this metabolism. To get a first clue to which genes are responsible for this metabolism I wanted to search these genomes for genes that are present only in one set of strains but absent in another set of strains.

Preferably, I would like to try to learn a programming language to figure this, and future questions like this, out by myself. I have started learning R using datacamp and coursera John Hopkins course on Next Generation Sequencing.

So what would be the best approach for this? Do you think this would be doable if I commit myself to learning R? Or will this take years to be able to solve? Also, for answering these type of questions, is it better to learn R or Python?

I would really appreciate your suggestions. Thanks in advance!

mastal 05-15-2016 11:39 AM

I think you will need both R and a scripting language (e.g. python), because they are good for different things.

westerman 05-16-2016 08:17 AM

I agree with Mastal: both are useful. What is even more useful is to learn hoe to use the existing libraries in both languages. Instead of writing ones own code learn how to find and use pre-existing code. For example I limp along in de novo R programming but knowing that Bioconductor exists I can do a lot in R.

ECO 05-16-2016 10:07 AM

Learning both is great. Python+Pandas is pretty powerful, but (IME) is harder to learn than R as the syntax of pands, and the intelligent default behaviors are not as robust as they are in R.

I do >50% of my smaller scale plotting/data munging in R (reading in flat files, etc), with ggplot2.

However anytime I want to interact with bam files, very large datasets, or create hundreds of plots, I turn to python. Plotting with pandas+seaborn is pretty nice.

Jessica_L 05-16-2016 10:48 AM

Seconding the comment about the utility of bioconductor.

wdecoster 05-17-2016 02:03 AM

I think the learning curve of python is less steep than R (I started with Python and know both reasonably well). Both languages have their (dis)advantages, but everything you can do in one language you can also do in the other.

All times are GMT -8. The time now is 10:16 PM.

Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2021, vBulletin Solutions, Inc.