A method, system, device and medium for detecting SNP-SNP interactions

By combining particle swarm optimization with active learning and elite preservation mechanisms, the problems of high time complexity and low optimization efficiency in SNP-SNP interaction detection are solved, achieving more efficient detection results.

CN116631505BActive Publication Date: 2026-07-24QUFU NORMAL UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN ยท China
Patent Type
Patents(China)
Current Assignee / Owner
QUFU NORMAL UNIV
Filing Date
2023-05-29
Publication Date
2026-07-24

AI Technical Summary

Technical Problem

Existing SNP-SNP interaction detection methods suffer from high time complexity and low optimization efficiency.

Method used

The particle swarm optimization algorithm is adopted, with mutual information as the objective function. Through global and local search, combined with active learning strategy and elite preservation mechanism, the SNP-SNP interaction detection method is optimized.

Benefits of technology

It improves the computational efficiency of SNP-SNP interaction detection, ensures that elite particles are not discarded during the iteration process, and quickly finds the global optimum.

โœฆ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116631505B_ABST
    Figure CN116631505B_ABST
Patent Text Reader

Abstract

The embodiment of the application provides a SNP-SNP interaction detection method, system, device and medium, relates to the particle swarm optimization technical field, and is used to solve the technical problems that the existing SNP-SNP interaction detection has high time complexity and low optimization efficiency. Including: taking mutual information as a target function, and storing one SNP in each component of all particles; the mutual information is used to represent the correlation between SNP interaction and disease phenotype; all particles are sequentially filled into a plurality of sub-populations from top to bottom in descending order; wherein, the number of particles stored in each sub-population is consistent; the plurality of sub-populations are globally and locally searched to complete algorithm iteration; after the algorithm iteration is terminated, the position of the particle with the highest target function value is obtained, and the correlation between the SNP-SNP interaction corresponding to the particle and the disease phenotype is the strongest.
Need to check novelty before this filing date? Find Prior Art