RNA Secondary Structure Prediction By Learning Unrolled Algorithms
This work addresses RNA structure prediction, a key problem in computational biology, by introducing a novel deep learning approach that improves accuracy for complex structures like pseudoknots, representing a strong specific gain in the domain.
The paper tackled RNA secondary structure prediction by proposing E2Efold, an end-to-end deep learning model that directly predicts base-pairing matrices using unrolled algorithms for constraint enforcement, resulting in significantly better structure predictions, especially for pseudoknotted structures, with inference efficiency comparable to the fastest algorithms.
In this paper, we propose an end-to-end deep learning model, called E2Efold, for RNA secondary structure prediction which can effectively take into account the inherent constraints in the problem. The key idea of E2Efold is to directly predict the RNA base-pairing matrix, and use an unrolled algorithm for constrained programming as the template for deep architectures to enforce constraints. With comprehensive experiments on benchmark datasets, we demonstrate the superior performance of E2Efold: it predicts significantly better structures compared to previous SOTA (especially for pseudoknotted structures), while being as efficient as the fastest algorithms in terms of inference time.