Protein Classes Predicted by Molecular Surface Chemical Features: Machine Learning-Assisted Classification of Cytosol and Secreted Proteins.

ABSTRACT: Chemical structures of protein surfaces govern intermolecular interaction, and protein functions include specific molecular recognition, transport, self-assembly, etc. Therefore, the relationship between the chemical structure and protein functions provides insights into the understanding of the mechanism underlying protein functions and developments of new biomaterials. In this study, we analyze protein surface features, including surface amino acid populations and secondary structure ratios, instead of entire sequences as input for the classifier, intending to provide deeper insights into the determination of protein classes (cytosol or secreted). We employed a random forest-based classifier for the prediction of protein locations. Our training and testing data sets consisting of secrete

SUBMITTER: Hu G 

PROVIDER: S-EPMC11382266 | biostudies-literature | 2024 Sep

REPOSITORIES: biostudies-literature