Protein Structure Prediction : A Primer

INTRODUCTION

Structure prediction 🧩 and protein folding 🧬always amazed me, not just from the perspective of a commoner drooling over its infinite potential in medical field, but, also as an engineer on how an amalgamation of multiple modalities and learning methods can pave way for such a beautiful framework. With major breakthroughs from Alphafold and Alphafold2 and open models like ESM and ESM2, the entire pathway is bustling with great insights and learning.🤖

I had been following this topic for quite some time now, but, I always lack that primer over why? and How? 🤔 of the entire problem of structure prediction itself. Hence, decided to share excerpts of what I learnt on this in a series of posts. Covering basics of structure prediction to deep learning based highly performant models.

Topics 📑
1. What is structure prediction? why is it helpful?
2. How Structure prediction is conventionally handled (all steps)? and Where does deep learning fit in bigger picture?
3. Evaluation metrics and CASP.
4. Components and priors for models like Alphafold.
5. Training procedure and how to interpret results.
6. Models without priors (ESM).
8. Framework and tools.
9. Perspective from Diffusion and similar models.
10. CRISPR and its potential

So, stick with me for next few weeks to dive into this vast ocean of explorations and AHA!! moments. 🧑‍💻🙆‍♂️