# RNA secondary structure prediction by conducting multi-class classifications

## TL;DR

This paper introduces a new method for predicting RNA secondary structures by using multi-class classification, avoiding complex post-processing steps.

## Contribution

The novel approach frames RNA secondary structure prediction as multi-class classification, eliminating the need for post-processing.

## Key findings

- The proposed method produces valid predictions without complex post-processing steps.
- Additional methods like data augmentation improve performance within RNA families.
- The method is model-agnostic and can be used with matrix predictions of structures.

## Abstract

Generating valid predictions of RNA secondary structures is challenging. Several deep learning methods have been developed for predicting RNA secondary structures. However, they commonly adopt post-processing steps to adjust the model output to produce valid predictions, which are complicated and could limit the performance. In this study, we propose a simple method by considering RNA secondary structure prediction as multiple multi-class classifications, which eliminates the need for those complicated post-processing steps. Then, we use this method to train and evaluate our model based on the attention mechanism and the convolutional neural network. Besides, we introduce two additional methods, including data augmentation to further improve the within-RNA-family performance and a method to alleviate the performance drop in the cross-RNA-family evaluation. In summary, we could produce valid predictions and achieve better performance without complex post-processing steps, and we show our additional methods are beneficial to the performance in within-RNA-family and cross-RNA-family evaluations.

•We propose a new method that considers RNA secondary structure prediction as multi-class classifications.•We show that our method eliminates the need for complex post-processing steps to product valid secondary structures.•The method is model-agnostic and could be used to train and evaluate models using matrix predictions of structures.

We propose a new method that considers RNA secondary structure prediction as multi-class classifications.

We show that our method eliminates the need for complex post-processing steps to product valid secondary structures.

The method is model-agnostic and could be used to train and evaluate models using matrix predictions of structures.