Please turn JavaScript on
Oxford Protein Informatics Group icon

Oxford Protein Informatics Group

Is this your feed? Claim it!

Publisher:  Unclaimed!
Message frequency:  0.19 / day

Message History

SAbDab has been around since 2013. Its now over 22,000 antibody structures have been searched and downloaded and searched again by a great many people with a great many browser tabs open, clicking, and scrolling, and clicking away.

Starting this week, they no longer have to.

SAbDab2 now has a public API. You could read all about it in t...


Read full story

For the past few months, I’ve been working on an AI agent for antibody optimisation. The idea is to let the agent use existing computational tools to identify developability issues, suggest modifications, and evaluate the resulting candidates.

Evaluating such an agent might sound straightforward: is the optimised antibody better than the starting point? But obviousl...


Read full story

Some recent cheminformatics papers have been using “win rate” (i.e., how often a model is the best across a set of tasks) to evaluate benchmarking performance. Although it can be a useful proxy for how good a model is, I have some gripes with this metric, particularly if it is presented without additional context.

Firstly, it can hide small differences; a model may...


Read full story

AKA why a protein does not naturally fit neatly into a tensor, and how RFdiffusion3, BoltzGen and La-Proteina cope when the amino-acid sequence is not known yet

The problem: proteins are not rectangles

Neural networks like fixed-shape arrays. Proteins are not fixed-shape: the 20 standard amino acids contain between 4 and 14 heavy atoms (hydrogens are usu...


Read full story

Andrew D. Blevins and Ian K. Quigley, co-founders of Leash Bio, released a pre-print last year that I think deserves some more attention. Using a Bemis-Murcko scaffold split, the authors show that it is possible to train a machine learning model to identify which of 1,815 medicinal chemists produced a molecule from its structure alone, achieving a top 5 accuracy of 60% and a ...


Read full story