Method for predicting phosphorylation site-specific kinase

Abstract

The invention relates to a method for predicting phosphorylation site-specific kinase, which comprises the following specific steps of: constructing a kinase-substrate heterogeneous network according to data in a database, and calculating different weights given to edges of the kinase-substrate heterogeneous network; on the basis of the established network, predicting a potential kinase-substrate relationship by adopting a weighted maximum bipartite graph matching algorithm; for the queried phosphorylation site, intercepting an amino acid fragment of the queried phosphorylation site, and calculating sequence similarity scores between the fragment and substrate fragments catalyzed by different kinases; and predicting the catalytic kinase of the given phosphorylation site by combining the predicted kinase-substrate relationship and sequence similarity score. According to the prediction method of the phosphorylation site specific kinase, PPI network information and protein local sequence information are effectively combined; catalytic possibilities of each kinase under different thresholds are given according to the query site; compared with the existing method, the method has higher prediction accuracy on different kinases.

Description

technical field [0001] The invention relates to the field of biological information, in particular to a method for predicting phosphorylation site-specific kinases. Background technique [0002] Phosphorylation catalyzed by protein kinases is the most ubiquitous post-translational modification in eukaryotes and plays an important regulatory role in almost all cellular behaviors, including DNA repair, transcription regulation, apoptosis, immune response, metabolism and cell differentiation . More and more experimental studies have revealed that protein kinases can directly or indirectly affect the abnormal phosphorylation of proteins, and are involved in the occurrence and development of many serious diseases. Therefore, the study of the specific effect of kinases on protein phosphorylation sites is of great significance for understanding the regulatory mechanism of kinases in physiological and pathological processes and the design of drugs for related diseases. [0003] Wi...

AI Technical Summary

Problems solved by technology

Although some computational tools have been proposed, the currently available prediction tools have the following shortcomings: some tools can only predict the relationship between kinases and substrate proteins, and cannot accurately predict the catalytic kinase at the phosphorylation site; most of the tools Based on the machine learning algorithm, it is necessary to encode the features of the samples. This process will inevitably cause the loss of original information. In addition, the unbalanced positive and negative sample sets will also affect the prediction results of machine learning; finally, some methods are used to extract PPI The network information only considers local information and does not use global information, which may lead to false positives

Images