AI Article Synopsis
- Predicted inter-residue distances play a crucial role in improving the accuracy of protein structure prediction (PSP), but combining short and long distance predictions remains difficult.
- A new stacked meta-ensemble method is introduced to effectively combine deep learning models for predicting various ranges of distances, demonstrating over a 5% improvement in Local Distance Difference Test (LDDT) scores compared to existing methods.
- The method, called meta-ensemble for distance prediction (MDP), also shows that using predicted long distances enhances protein conformations more than relying solely on predicted short distances, with the program available online.
Article Abstract
Predicted inter-residue distances are a key behind recent success in high quality protein structure prediction (PSP). However, prediction of both short and long distance values together is challenging. Consequently, predicted short distances are mostly used by existing PSP methods. In this paper, we use a stacked meta-ensemble method to combine deep learning models trained for different ranges of real-valued distances. On five benchmark sets of proteins, our proposed inter-residue distance prediction method improves mean Local Distance Different Test (LDDT) scores at least by 5% over existing such methods. Moreover, using a real-valued distance based conformational search algorithm, we also show that predicted long distances help obtain significantly better protein conformations than when only predicted short distances are used. Our method is named meta-ensemble for distance prediction (MDP) and its program is available from https://gitlab.com/mahnewton/mdp.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1016/j.compbiomed.2022.105824 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Want AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!