Towards a robust out-of-the-box neural network model for genomic data

TL;DR

This paper evaluates the robustness and transferability of neural network models for genomic data, highlighting recurrent neural networks' superior performance and identifying baseline characteristics for future model development.

Contribution

It provides an analysis of neural network robustness in genomics, emphasizing recurrent models and proposing baseline features for out-of-the-box applications.

Findings

Recurrent neural networks outperform convolutional models in accuracy and transferability.

Neural networks face challenges due to biological data heterogeneity and modest sample sizes.

Certain model characteristics are identified as transferable across datasets.

Abstract

The accurate prediction of biological features from genomic data is paramount for precision medicine and sustainable agriculture. For decades, neural network models have been widely popular in fields like computer vision, astrophysics and targeted marketing given their prediction accuracy and their robust performance under big data settings. Yet neural network models have not made a successful transition into the medical and biological world due to the ubiquitous characteristics of biological data such as modest sample sizes, sparsity, and extreme heterogeneity. Here, we investigate the robustness, generalization potential and prediction accuracy of widely used convolutional neural network and natural language processing models with a variety of heterogeneous genomic datasets. Mainly, recurrent neural network models outperform convolutional neural network models in terms of prediction accuracy, overfitting and transferability across the datasets under study. While the perspective of a robust out-of-the-box neural network model is out of reach, we identify certain model characteristics that translate well across datasets and could serve as a baseline model for translational researchers.