A survey of how to use protein language models for protein design: Part 3

Rounding out the series on PLMs... and a plug for a job posting at Moderna!

Hello fellow datanistas!

Have you ever wondered how to effectively evaluate protein language models (PLMs) and prioritize which generated sequences to test in the lab? If so, I've just published the third part of my series on protein language models that dives deep into this very topic. This latest installment focuses on the critical evaluations of PLMs, a cornerstone for anyone involved in computational biology and protein engineering.

In this post, I explore the various types of evaluations that can be applied to generated protein sequences, from sequence-based to structure-based evals, and discuss how these can significantly influence the decision-making process in protein design. Whether you're a lab scientist or a computational biologist, understanding these evals is crucial for optimizing the testing of generated sequences and ensuring that your efforts yield functional and effective proteins.

I also share insights on how computational and laboratory teams can collaborate more effectively, ensuring that both sides can leverage their strengths to enhance the outcomes of protein design projects. This is a must-read for anyone looking to bridge the gap between computational predictions and laboratory validations.

I invite you to read the full post here and join me in this fascinating exploration of protein language models. If you find the post enlightening, please consider forwarding it to others in your network who might benefit from these insights. Together, we can push the boundaries of what's possible in protein engineering.

Thank you for your continued support and curiosity!

Happy coding!
Eric

P.S. making a plug for my colleague Anand Murthy, who is hiring a software developer to work on the Machine Learning Platform at Moderna. He is based in Seattle, but works very, very closely with the data science teams in Cambridge, which includes myself. If you’re interested, please check out the job posting here!