Predication method for biological function of GPCR (G Protein-Coupled Receptor) drug target and application of predication method

Abstract

The invention provides a predication method for the biological function of a GPCR (G Protein-Coupled Receptor) drug target and application of the predication method. The predication method comprises the following steps: firstly acquiring feature information of a GPCR, using the feature information to form a corresponding view; evaluating the performance of the feature view of the GPCR through an inductive matrix complementing method, finding out the feature information effective for the construction of a model; complementing a potential GPCR-GO relation besides the known GPCR-GO relation matrix, and then searching a part having a real relationship in Pij=0 to predicate the biological function of the GPCR. The predication method is based on a multi-view inductive matrix complementing (IMC)algorithm, and the feature of a marker is also added while the feature of the GPCR of a sample is taken into consideration. In addition, the experiment result shows that compared with the traditionalmatrix complementing algorithm, the performances are obviously improved. The problem of serious class imbalance of the multi-marker algorithm in the prior art is solved.

Description

technical field [0001] The invention relates to the technical field of machine learning prediction, in particular to a method for predicting the biological function of a GPCR drug target and its application. Background technique [0002] G protein-coupled receptor (GPCR) is a seven-transmembrane domain receptor, which is an important protein in cell signal transduction. all play an important role. When bound to an appropriate ligand outside the cell membrane, activated G proteins can be used to transduce signals outside the cell. G protein-coupled receptors (GPCRs) are one of the target enzymes and are the targets of many drugs. The key to modern new drug research and development is to find, determine and prepare drug screening targets-molecular drug targets. Drug targets refer to the binding sites of drugs in the body, including gene sites, receptors, enzymes, ion channels, nucleic acids and other biomacromolecules. Determining new effective drug targets is the primary ...

AI Technical Summary

Problems solved by technology

However, there is often a certain amount of information redundancy in different view data. How to fully extract useful features from multi-view data and eliminate redundancy has become a key issue in multi-view learning technology.

Images