If you want to do single or multi-mutant protein optimization with a (very) small amount of data and a laptop, check out our recent work FolDE (github.com/JBEI/foldy)! Open platform with a super nice, intuitive UI! Led by @jacoberts.bsky.social at @keaslinglab.bsky.social.

We're excited to share FolDE, a low-N protein optimization method. In simulation, we found that FolDE is 55% more likely to identify top-1% hits than current baseline methods. FolDE is open and can be set up on a personal computer with a single command. arxiv.org/abs/2510.24053