Thank you for the advice. That makes a lot of sense it's easy to get caught up trying to learn every tool and concept before actually starting. I like the idea of learning through a real project and picking up the necessary skills along the way. As someone still getting started, do you have any suggestions for beginner-friendly projects or biological questions that would be good to work on first?
I am new to computational biology and bioinformatics and would appreciate some guidance from people already working in the field. I am currently studying pharmacy. While learning about drug discovery, I became interested in computer-aided drug design (CADD) and realized that many modern drug discovery approaches rely on computational biology and bioinformatics. This motivated me to start learning programming and computational methods, but I am unsure how to build a solid foundation in the field. I have recently started learning Python and basic biological sequence analysis using Biopython.
I am feeling a bit overwhelmed by the number of topics in the field—genomics, transcriptomics, machine learning, structural biology, statistics, workflow management, and many others. I would like to build a strong foundation but am unsure what skills I should prioritize first.
My questions are:
1.What core subjects should a beginner focus on during the first 6–12 months? 2.How much programming, statistics, and molecular biology knowledge is essential at the beginning? 3.Are there any books, courses, or practical projects that you would recommend? 4.Looking back, what do you wish you had learned earlier when starting in computational biology?
I would be grateful for any advice, learning roadmaps, or personal experiences.
Thank you!
1 answer
My suggestion for this kind of thing is always: learn by doing.
Pick up a project that you want to get stuck into. Solve a biological question (e.g. can I design a drug for this target?).
Worry less about "completing" stats, or a workflow tool intro, and just get stuck in. Initially, you'll hone your basic scripting and coding skills. As you get more accomplished you'll realise you're running the same analyses a lot potentially, and maybe its time to experiment with a workflow tool. You'll pull in different tools and analyses to answer different questions, and you'll get to grips with the various options as you experiment. This will require you to research what are the best tools for each task, meaning your familiarity with the software ecosystem and important literature grows.
All of these things feed off of one another.
Take a look at ROSALIND and work through the problems: https://rosalind.info/problems/locations/
Thanks for this !! I’ll check out ROSALIND and start working through the problems.
ROSALIND is good, but they're just toys.
I would suggest you start by trying to replicate an analysis in a paper or something if you aren't sure what project of your own you might want to do.
Take a look at some QSAR papers or something like that and see if you can run the analyses yourself. Wheverer you get stuck you know is a gap in your knowledge and you'll need to spend time working on that peice.
This.
I can’t seem to find it, but Ming Tang has a blog post (or maybe it was a YouTube video) demonstrating this kind of repro work step-by-step for beginners with ChIP-seq data. Could be useful for OP (ChibiOrbit) to check that stuff. (OP: I’d look here and here.)
Thank you for the suggestion! I'll see if I can find that blog post or video.
Thanks a lot for this advice. I'll definitely look into replicating a few papers and use the roadblocks as a roadmap for what I need to learn next.I appreciate you pointing me in that direction.
Log in to answer this question.
If you want to try basic bioinformatics and Linux tools and commands, check https://sandbox.bio. Additional resources https://github.com/danielecook/Awesome-Bioinformatics.