Gian Marco and Will interviewed by preLights

Our MULTI-evolve reanalysis was highlighted by preLights

Gian Marco and Will spoke with Leonie Brüne for a preLights spotLights episode about our preprint Additive baselines furnish no evidence for epistasis learning by MULTI-evolve. We talk about how the work came together, why additive models are the key baseline when a paper claims to learn epistasis, and directions in machine learning for protein engineering.

Note

For the backstory on the preprint see earlier post, Additivity is all you need?.