Unsupervised cross-domain translation via deep learning and adversarial attention neural networks and application to music-inspired protein designs

Taking inspiration from nature about how to design materials has been a fruitful approach, used by humans for millennia. In this paper we report a method that allows us to discover how patterns in disparate domains can be reversibly related using a computationally rigorous approach, the AttentionCrossTranslation model. The algorithm discovers cycle-and self-consistent relationships and offers a bidirectional translation of in-formation across disparate knowledge domains. The approach is validated with a set of known translation problems, and then used to discover a mapping between musical data-based on the corpus of note se-quences in J.S. Bach's Goldberg Variations created in 1741-and protein sequence data-information sampled more recently. Using protein folding algorithms, 3D structures of the predicted protein sequences are generated, and their stability is validated using explicit solvent molecular dynamics. Musical scores generated from protein sequences are sonified and rendered into audible sound.

Automated summary

Taking inspiration from nature, humans have used the design approach of materials for centuries. This paper introduces the AttentionCrossTranslation model, which computationally discovers reversible relationships among patterns in different domains. The model is validated with known translation problems and then used to map musical data to protein sequences. The generated protein structures are validated using molecular dynamics and the music scores are turned into audible sound.