DNA sequence cluster of Locality Sensitive Hashing based on standard entropy

Abstract

The invention discloses a DNA sequence cluster of a Locality Sensitive Hashing based on standard entropy. The method comprises the steps of mapping an original DNA sequence according to an L-Gram model, calculating a matrix constituted by LF entropy values of N pieces of sequences to obtain the standard entropy of the matrix, using the Locality Sensitive Hashing to conduct Hashing mapping on the standard entropy to obtain a candidate set of DNA fragment sequences, and calculating DNA fragment sequences of which the editing distance is shorter than d in the candidate set to obtain a cluster result. According to the DNA sequence cluster of the Locality Sensitive Hashing based on the standard entropy, the situation that enough original DNA information is included in a converted characteristic space is taken into full consideration, missing of the DNA information is avoided, each fragment of the DNA sequences is turned into a new space, a candidate DNA fragment sequence set of each piece of DNA fragment sequences is calculated, and thus the operation speed can be increased, and the operation precision can be improved.

Description

technical field [0001] The invention relates to the field of biological information processing, in particular to DNA sequence clustering based on local sensitive hashing of standard entropy. Background technique [0002] With the advent of the Internet age and the development of information technology, the development of gene sequencing technology has become more and more mature. In addition to the development of various genetic projects, the amount of biological data has increased exponentially. Traditional methods have been unable to meet the massive data processing and analysis. . Bioinformatics refers to the combination of biology and computer technology, interaction with mathematics disciplines, acquisition of biological information for its processing, extraction, analysis, storage, etc., and mining of genetic material location information. Data mining technology is a technology that can extract useful and potentially effective information from a large amount of data. ...

AI Technical Summary

Problems solved by technology

Based on the division-based K-medoid algorithm and the hierarchical-based complete-link algorithm, these traditional clustering algorithms require pairwise comparison of sequences, which has a high time complexity. Today, the number of DNA sequences is growing extremely fast. Algorithms cannot be applied to massive data

The K-means algorithm needs to determine the number of clusters, and the centroid of sequence data is not easy to calculate. The random initial cluster center makes the clustering result unstable, and the clustering effect when applied to biological sequence data is not good.

The result of the clustering algorithm based on the BAG graph is effective, but it needs to be guided by the clustering unit when dividing the class, and the number of sequences in the gene pool is too large, which makes it extremely difficult to use an undirected graph to represent too many sequences

Images