ProQ3D: Improved model quality assessments using Deep Learning

TL;DR

ProQ3D introduces a deep learning approach to protein quality assessment, significantly enhancing correlation metrics over previous models by replacing SVMs with neural networks while using the same input features.

Contribution

The paper demonstrates that replacing SVMs with deep neural networks on existing features markedly improves protein quality assessment accuracy.

Findings

Pearson correlation improved to 0.90 with ProQ3D.

Same input features as ProQ2/ProQ3 yield better results with deep learning.

ProQ3D is freely available as a webserver and stand-alone tool.

Abstract

Summary: Protein quality assessment is a long-standing problem in bioinformatics. For more than a decade we have developed state-of-art predictors by carefully selecting and optimising inputs to a machine learning method. The correlation has increased from 0.60 in ProQ to 0.81 in ProQ2 and 0.85 in ProQ3 mainly by adding a large set of carefully tuned descriptions of a protein. Here, we show that a substantial improvement can be obtained using exactly the same inputs as in ProQ2 or ProQ3 but replacing the support vector machine by a deep neural network. This improves the Pearson correlation to 0.90 (0.85 using ProQ2 input features). Availability: ProQ3D is freely available both as a webserver and a stand-alone program at http://proq3.bioinfo.se/