Why and how should you start learning bioinformatics?

Learning bioinformatics in a self-directed way might feel like getting lost in a forest but there is plenty of information online – and support from others – that can help guide your learning. On this useful post, Fernando Pozo suggests approaches and resources to start your bioinformatics journey.
Why and how should you start learning bioinformatics?

TL;DR - My 1st June, 2022 post on Twitter covers some of the most important tips described here.

Spoiler alert: After reading this article, you will have full access to most of the materials you need to start managing bioinformatics. Hold your money, book a time to enjoy learning, open your laptop and please click all the URLs!

Why should you learn bioinformatics?

2020, December: John Moult, a computational biologist who co-founded CASP (the Critical Assessment of Structure Prediction) more than 25 years ago, declared that the single-chain protein structure prediction was solved. DeepMind, Google's AI research group led by Demis Hassabis, was the team responsible for creating this breakthrough. It possibly represents the first time computer science and artificial intelligence have significantly advanced the frontiers of humanity's scientific knowledge. It will dramatically solve many 21st-century science problems involving living systems. Now in 2022, Hassabis has been recently honoured with the prestigious Spain's Princess of Asturias Award (among many other recognitions), and bioinformatics, the scientific field surrounded by this discovery, is one of the trendiest disciplines in life sciences.

Watch: Protein folding explained is less than 2 minutes

The National Human Genome Research Institute defines bioinformatics as the scientific discipline that intends to collect, store, analyse and disseminate biological data and information from the genome, such as DNA and amino acid sequences or annotations about those sequences. The collected data and information will be used afterwards to provide insights into life sciences. This definition leads any researcher to a fantastic but, at the same time, gigantic stage.

How should you learn bioinformatics?

At this point, if you want to introduce, learn, or even gain expertise in bioinformatics: how should you start learning? Please, continue reading to see my 10 noise-free recommendations.

  1. Independently of your pre-obtained knowledge in bioinformatics, start discovering free and curated resources like the ones I will show here. The traditional education environment in this area is full of outdated and monotonous courses that will not assist with real projects.
  2. The ability to manage freely available resources is potentially powerful. Firstly, you will learn with significantly updated resources (and as you have noticed in the first paragraph, this is crucial here) and, secondly, taking the opportunity to collaborate or interact with content creators worldwide is precious.
  3. Follow a specific and concrete training plan. People don't want to swim in a universe of Coursera-like MOOCs, which are often not concrete or updated (and can be expensive). Even the university course selection can be frustrating (and often doesn't follow the previous premises). Consider skipping more general courses.
  4. Foundations in bioinformatics are covered here by Professor Aaron Quinlan in the Applied Computational Genomics course. (23 videos of 60'-80' each one) with sorted and updated lectures.
  5. Unix-like OS (like Ubuntu or MacOS) are your best friends to become a bioinformatician. Missing Semester shows how you use it in 11 videos of 40'-60' each.
  6. Learning programming is going to be a must. Start with a concrete and user-friendly language (Python is my recommendation), and you will be ready to discover new things. The Carpentries have lots of resources prepared for you. If you feel ready and enjoy programming, hypermodern Python techniques will provide you with the rest.
  7. AlphaFold2 used Machine Learning and Deep Learning modern techniques as part of their significant breakthrough. This field is vast, but if you want to understand the intersection with bioinformatics, in MIT Deep Learning in Life Sciences, Professor Manolis Kellis provides a very detailed course (22 videos of 60'-120').
  8. If you feel ready to understand most of the previous concepts, the fantastic community of nf-core has already set tons of pipelines (entirely ready for production) for you. Among +50 pipes, RNA-seq, variant calling, ChIP-seq, or ATAC-seq end-to-end pipelines are curated, accessible, free, and prepared for your scientific data.
  9. Finally, what's happening with interacting with other people in the field? It is always a crucial step in learning! Bioinformaticians are over Discord, Reddit, nf-core slack, or RSG countries communities (check for the community created in Spain). GitHub joins +83M of software developers and therefore, is another wonderful community of bioinformatics scientists, and, by the way, this is a different and awesome list of bioinformatics resources.
  10. Extra tip: Still not curious about the power of bioinformatics? Let's start reading one of these three free and inspiring resources inside a traditional book: Modern Statistics for Modern Biology, Cell Biology by the Numbers, Interpretable Machine Learning.

I want to thank the people who take care of creating all this fantastic content accessible to everyone entirely for free. I wish you a good bioinformatics trip!

Fernando Pozo


Photo by Caleb Jones on Unsplash 

Comments

Go to the profile of João Victor Cabral Costa
about 1 month ago

Great resources, @Fernando Pozo. Thanks for putting it together!