Recent machine learning-based methods - including RFdiffusion, FoldFlow-2 and BindCraft - now enable the de novo design of synthetic proteins that interact with any specific protein, nucleic acid, or small molecule of interest. For example, synthetic miniproteins are small, ultrastable structures that can recognize target proteins with affinities in the nM-pM range. De novo designed synthetic proteins, scaffolds, antibodies and enzymes will have broad applications as research tools, diagnostics, and therapeutics.

This BioGRID themed curation project on synthetic protein interactions aims to annotate de novo designed proteins that have been experimentally validated to interact with the intended target. This curation project will be updated each month and expanded as new design classes are developed. As synthetic proteins are not currently annotated in standard protein resources, each entry is assigned a unique BioGRID identifier. If you notice errors or missing interactions/publications in this dataset, or otherwise wish to contribute to this project, please contact us at support@thebiogrid.org.