SAAMBE-SEQ is a sequence-based machine learning algorithm to predict the binding energy changes upon single mutation in protein-protein complexes. Unlike other existing methods, SAAMBE-SEQ does not require a 3D complex structure as input. This method can be applied to protein complexes without known structure. The accuracy of SAAMBE-SEQ is comparable to existing structure-based methods. This method can be used to provide understanding the effect of mutation in protein complexes using sequence alone.
Reference: Gen Li, Swagata Pahari, Adithya Krishna Murthy, Siqi Liang, Robert Fragoza, Haiyuan Yu, Emil Alexov, SAAMBE-SEQ: A Sequence-based Method for Predicting Mutation Effect on Protein-protein Binding Affinity, Bioinformatics, btaa761. available here