A novel sequence alignment algorithm based on deep learning of the protein folding code

Motivation: From evolutionary interference, function annotation to structural prediction, protein sequence comparison has provided crucial biological insights. While many sequence alignment algorithms have been developed, existing approaches often cannot detect hidden structural relationships in the 'twilight zone' of low sequence identity. To address this critical problem, we introduce a computational algorithm that performs protein Sequence Alignments from deep-Learning of Structural Alignments (SAdLSA, silent 'd'). The key idea is to implicitly learn the protein folding code from many thousands of structural alignments using experimentally determined protein structures. Results: To demonstrate that the folding code was learned, we first show that SAdLSA trained on pure alpha-helical proteins successfully recognizes pairs of structurally related pure beta-sheet protein domains. Subsequent training and benchmarking on larger, highly challenging datasets show significant improvement over established approaches. For challenging cases, SAdLSA is similar to 150% better than HHsearch for generating pairwise alignments and similar to 50% better for identifying the proteins with the best alignments in a sequence library. The time complexity of SAdLSA is O(N) thanks to GPU acceleration.