Intelligent forecast model construction technology of fist class protein secondary structure

Abstract

The invention discloses a technology for constructing an intelligent prediction model of a protein secondary structure, and the model is integrated by a structural model with multi-layer recursive and stepwise refinement. The model CPM fuses an inventive KAAPRO method, a novel homologous analysis method, an improved SVM method and the like. The CPM breaks through the technical route of the traditional single physical and chemical attribute analysis or a single structural sequence analysis, and adopts a preferred route by combining the structural sequence analysis and the physical and chemical attribute analysis to ensure the optimization of the whole model good prediction precision and better universality. The CPM is mined by a high-starting-point alpha/beta base and goes through domain knowledge and background knowledge; and the CPM can better predict the secondary structure of a meta-alpha/beta protein and has the highest precision of 86% (the highest precision of the similar type is 81%).

Description

technical field [0001] The present invention relates to protein secondary structure prediction technology, in particular to a kind of intelligent prediction technology based on synthetic pyramid model (CPM, Compound Pyramid Model), specifically a kind of brand-new progressive refinement, multi-layer hierarchical protein secondary structure Structural Predictive Models - Synthetic Pyramid Model Construction Techniques. Background technique [0002] 1. Protein structure prediction technology: [0003] Protein structure prediction is an important task in the post-genomic era. Since the end of the 20th century, the field of molecular biology and bioinformatics (Bioinformatics) has made great progress. The Human Genome Project has basically completed the determination of the human gene sequence, and then entered the post-genome era. Since it is generally believed that the structure of a protein determines its function, and the rapid growth of protein sequence data has continuou...

AI Technical Summary

Problems solved by technology

At present, most KDD algorithms do not study KDD as a complex system of cognition and its internal regularity, and do not consider the knowledge base in depth. Duplicated and redundant, even incompatible, and relying only on human-computer interaction to form a focus, without reflecting the cognitive autonomy of the system itself, so the novelty and effectiveness required in the definition of KDD cannot be reflected

Images