Abstract
Several studies have linked disruptions of protein stability and its normal functions to disease. Therefore, during the last few decades, many tools have been developed to predict the free energy changes upon protein residue variations. Most of these methods require both sequence and structure information to obtain reliable predictions. However, the lower number of protein structures available with respect to their sequences, due to experimental issues, drastically limits the application of these tools. In addition, current methodologies ignore the antisymmetric property characterizing the thermodynamics of the protein stability: a variation from wild-type to a mutated form of the protein structure () and its reverse process () must have opposite values of the free energy difference (). Here we propose ACDC-NN-Seq, a deep neural network system that exploits the sequence information and is able to incorporate into its architecture the antisymmetry property. To our knowledge, this is the first convolutional neural network to predict protein stability changes relying solely on the protein sequence. We show that ACDC-NN-Seq compares favorably with the existing sequence-based methods.
Highlights
Received: 14 May 2021Accepted: 9 June 2021Published: 12 June 2021Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.Licensee MDPI, Basel, Switzerland.Predicting protein stability changes upon genetic variations is still an open challenge.It is essential to understand the impact of the alterations in the amino acid sequence, mainly due to non-synonymous DNA variations leading to the disruption or the enhancement of the protein activity, on human health and disease [1,2,3,4]
The protein stability changes upon variations of the amino acid sequence is usually expressed as the Gibbs free energy of unfolding (∆∆G), which is defined as the difference between the energy of the mutated structure of the protein and its wild-type form
Since the application of a deep learning technique requires a large amount of data to achieve the best performance, we pre-trained ACDC-NN-Seq using the predictions of another method, DDGun3D, which has shown to achieve antisymmetry with good performance
Summary
Received: 14 May 2021Accepted: 9 June 2021Published: 12 June 2021Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.Licensee MDPI, Basel, Switzerland.Predicting protein stability changes upon genetic variations is still an open challenge.It is essential to understand the impact of the alterations in the amino acid sequence, mainly due to non-synonymous (or missense) DNA variations leading to the disruption or the enhancement of the protein activity, on human health and disease [1,2,3,4]. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Predicting protein stability changes upon genetic variations is still an open challenge. It is essential to understand the impact of the alterations in the amino acid sequence, mainly due to non-synonymous (or missense) DNA variations leading to the disruption or the enhancement of the protein activity, on human health and disease [1,2,3,4]. Protein stability perturbations have already been associated to pathogenic missense variants [5]. The protein stability changes upon variations of the amino acid sequence is usually expressed as the Gibbs free energy of unfolding (∆∆G), which is defined as the difference between the energy of the mutated structure of the protein and its wild-type form
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
