Opinion
What we got wrong about affinity
For three years we optimised how tightly our designs bind. Specificity was the number that actually mattered.
It is easy to make a protein bind more tightly and hard to make it bind only one thing. We spent too long on the easy problem.
The turning point
A design with the best affinity we had ever measured also bound four unintended sites. A weaker design bound one.
What changed
We now report a specificity ratio next to every affinity, and we select on the ratio. Our review article makes the longer argument.
← Previous post