Getting Started With Ngs Analysis For Modeling Guy

Question

[Pappu]

I have some experience with molecular modeling, computational chemistry, bioinformatics especially python programming. Analysing NGS data seems a very promising field for me because amount of data is growing rapidly and a lot of interesting results can be obtained by 1000 genomes data.

However I am not sure how to get started with analysis such data. My long term goal is to incorporate modeling results into the NGS sequence analysis. I noticed that I have already used many of the statistical tools like clustering, PCA, network building, statistical significance tests for my work. So I will appreciate if someone can tell me how to proceed and what are the things that can be done with such data. Thank you.

Answer 1

[Zev.Kronenberg]

Just start reading. I know that sound pretentious, but it is really the best way to get going. Find a few papers in your field and read the methods carefully. If you have programming experience NGS analysis will be fun. Also definitely do NOT try to reinvent the wheel. There are a lot of tools out there already, using them will save you time so you can focus on novel problems.

Answer 2

[Dan D]

I came back into the bioinformatics field after a stint doing corporate programming. One high-level approach that helped me get up-to-speed with current analysis techniques was to do the tutorials at usegalaxy.org. I'm not necessarily advocating the use of Galaxy for your subsequent work, but it's a web-based interface to plenty of the most commonly-used tools. That way you can get an idea for how things work without investing too much energy or time. The tutorials are very well-done and informative and are based on analyses that an investigator would actually want to do in a real-world situation. Then you can dig deeper after you identify the particular tools in the tutorial pipelines that are of interest to you.

And like Zev mentioned, definitely read the papers. Nature Methods has been good about publishing NGS "how-to" papers, and most of the popular NGS tools will have at least one associated journal article.