CrMP-Sol database: classification, bioinformatic analyses and comparison of cancer-related membrane proteins and their water-soluble variant designs

BMC Bioinformatics

Research

Shuguang Zhang¹  Rui Qing²  Wenting Huang²  Ping Xu²  Sitao Zhang²  Hui Wang²  Lina Ma²  Fei Tao²  Emily Pan³  Jin Tang⁴  Qi Liang⁴ 

[1] Media Lab, Massachusetts Institute of Technology, 77 Massachusetts Avenue, 02139, Cambridge, MA, USA;State Key Laboratory of Microbial Metabolism, School of Life Sciences and Biotechnology, Shanghai Jiao Tong University, 200240, Shanghai, China;The Lawrenceville School, 2500 Main Street, 08648, Lawrenceville, NJ, USA;Zhejiang Lab, Research Center for Intelligent Computing Platforms, 311121, Hangzhou, Zhejiang, China;

关键词: Membrane protein;    Protein design;    QTY code;    Machine learning;    Protein function;    Cancer;    Bioinformatics;   

DOI  :  10.1186/s12859-023-05477-9

 received in 2023-01-25, accepted in 2023-09-12,  发布年份 2023

来源: Springer

【 摘 要 】

Membrane proteins are critical mediators for tumor progression and present enormous therapeutic potentials. Although gene profiling can identify their cancer-specific signatures, systematic correlations between protein functions and tumor-related mechanisms are still unclear. We present here the CrMP-Sol database (https://bio-gateway.aigene.org.cn/g/CrMP), which aims to breach the gap between the two. Machine learning was used to extract key functional descriptions for protein visualization in the 3D-space, where spatial distributions provide function-based predictive connections between proteins and cancer types. CrMP-Sol also presents QTY-enabled water-soluble designs to facilitate native membrane protein studies despite natural hydrophobicity. Five examples with varying transmembrane helices in different categories were used to demonstrate the feasibility. Native and redesigned proteins exhibited highly similar characteristics, predicted structures and binding pockets, and slightly different docking poses against known ligands, although task-specific designs are still required for proteins more susceptible to internal hydrogen bond formations. The database can accelerate therapeutic developments and biotechnological applications of cancer-related membrane proteins.

【 授权许可】

CC BY   
© BioMed Central Ltd., part of Springer Nature 2023

【 图 表 】

Fig. 1

Fig. 3

Fig. 3

Fig. 2

Fig. 7

12888_2023_5172_Article_IEq34.gif

Fig. 5