Predicting protein-peptide binding sites with a deep convolutional neural network.

J Theor Biol

Institute for Integrated and Intelligent Systems, Griffith University, Queensland, Australia; Laboratory for Medical Science Mathematics, RIKEN Center for Integrative Medical Sciences, Yokohama, Japan; CREST, JST, Tokyo 113-8510, Japan; School of Engineering and Physics, The University of the South Pacific, Suva, Fiji. Electronic address:

Published: July 2020

Motivation: Interactions between proteins and peptides influence biological functions. Predicting such bio-molecular interactions can lead to faster disease prevention and help in drug discovery. Experimental methods for determining protein-peptide binding sites are costly and time-consuming. Therefore, computational methods have become prevalent. However, existing models show extremely low detection rates of actual peptide binding sites in proteins. To address this problem, we employed a two-stage technique - first, we extracted the relevant features from protein sequences and transformed them into images applying a novel method and then, we applied a convolutional neural network to identify the peptide binding sites in proteins.

Results: We found that our approach achieves 67% sensitivity or recall (true positive rate) surpassing existing methods by over 35%.

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.jtbi.2020.110278DOI Listing

Publication Analysis

Top Keywords

binding sites
16
protein-peptide binding
8
convolutional neural
8
neural network
8
peptide binding
8
predicting protein-peptide
4
binding
4
sites
4
sites deep
4
deep convolutional
4

Similar Publications

To address the challenge of antibiotic-containing wastewater, a novel micromagnetic carrier-modified integrated fixed-film activated sludge system (MC-IFAS) was developed for treating tetracycline (TC)-containing swine wastewater in this study. The magnetic effects of the MC significantly enhanced TC removal by improving TC biosorption and biodegradation in both the suspended activated sludge and the carrier-attached biofilm in the MC-IFAS. The increased electrostatic attraction and number of binding sites in both the activated sludge and the biofilm enhanced their TC biosorption capacities, particularly in the activated sludge.

View Article and Find Full Text PDF

Palladium (Pd) catalysts are promising for electrochemical reduction of CO to CO but often can be deactivated by poisoning owing to the strong affinity of *CO on Pd sites. Theoretical investigations reveal that different configurations of *CO endow specific adsorption energies, thereby dictating the final performances. Here, a regulatory strategy toward *CO absorption configurations is proposed to alleviate CO poisoning by simultaneously incorporating Cu and Zn atoms into ultrathin Pd nanosheets (NSs).

View Article and Find Full Text PDF

Pancreatic ductal adenocarcinoma (PDAC) is highly aggressive and lacks effective therapeutic options. Cancer cells frequently become more dependent on splicing factors than normal cells due to increased rates of transcription. Terminal uridylyltransferase 1 (TUT1) is a specific terminal uridylyltransferase for U6 small nuclear RNA (snRNA), which plays a catalytic role in the spliceosome.

View Article and Find Full Text PDF

Immunologic bile duct destruction is a pathogenic condition associated with vanishing bile duct syndrome (VBDS) after liver transplantation and hematopoietic stem-cell transplantation. As the bile acid receptor sphingosine 1-phosphate receptor 2 (S1PR2) plays a critical role in recruitment of bone marrow-derived monocytes/macrophages to sites of cholestatic liver injury, S1PR2 expression was examined using cultured macrophages and patient tissues. Bile canaliculi destruction precedes intrahepatic ductopenia; therefore, we focused on hepatocyte S1PR2 and the downstream RhoA/Rho kinase 1 (ROCK1) signaling pathway and bile canaliculi alterations using three-dimensional hepatocyte culture models that form obvious bile canaliculus-like networks.

View Article and Find Full Text PDF

Phytoplankton plays a crucial role in the fate of pollutants in aquatic ecosystems by biotransformation and bioaccumulation. Aniline was listed in priority pollutants due to its toxicity and widespread distribution in the aquatic environment. This study focused on investigating the capacity and mechanism of eukaryotic alga Chlamydomonas reinhardtii in transforming aniline.

View Article and Find Full Text PDF

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!