Quantitative Structure-activity Relationship Model of Anti-rheumatoid Immune-active Peptides Based on Partial Least Square Method
Article
Figures
Metrics
Preview PDF
Reference
Related
Cited by
Materials
Abstract:
In this study, partial least squares (PLS) method was used to build a quantitative structure-activity relationship (QSAR) model for anti-rheumatoid immune activity of peptides. First, based on the bioactivity-related non-bonding effects (such as electrostatic, steric, and hydrophobic interactions) of each atom, the amino acids descriptor best suited for QSAR model of immune activity of peptides was selected to construct a QSAR model for a 47-peptide data set. The results showed that Z-scales amino descriptor achieved multiple-position and multivariate quantitative structure description for amino acid residues and was the most suitable for describing physicochemical properties of anti-rheumatoid immune-active peptides. Using the Z-scales descriptor, the model built on principle component analysis and PLS showed good reliability and predictive ability, with multiple correlation coefficient of 0.986, root mean square error (RMSE) of 0.253, leave-one-out cross-validation correlation coefficient of 0.893, and external validation coefficient of 0.971. Based on the structure-activity relationship of immune-active peptides, this study provides new opportunities for screening and design of new tolerogens for rheumatoid arthritis, thus providing a new approach for the development innovative functional food products.
