Protein anomaly detection method based on super-large scale evolutionary algorithm and hardware acceleration

Abstract

The invention discloses a protein anomaly detection method based on a super-large scale evolutionary algorithm and hardware acceleration. The method comprises the following steps: 1, acquiring mass spectrum characteristic data; 2, generating a mass spectrum characteristic selection scheme population and an external file, and setting parameters; 3, updating the external archive and quickly clustering and grouping the mass spectrum characteristics; 4, after mating pool selection is executed, a filial generation mass spectrum feature selection scheme is generated in the original space and the grouped reduced space at the same time; and 5, adaptively adjusting algorithm parameters by using a filial generation feature selection scheme, combining filial generation and parent generation mass spectrum feature selection scheme populations to perform environment selection, iterating the process to select a high-quality feature selection scheme, and finally obtaining an optimal protein anomaly detection feature selection scheme. According to the method, the problem of super-large-scale protein anomaly detection can be efficiently solved, a high-quality feature selection scheme is quickly obtained in a super-large-scale mass spectrum feature set, and protein anomaly detection is carried out, so that the detection accuracy is improved.

Description

technical field [0001] The invention relates to the technical field of feature selection for ultra-large-scale multi-objective problems, in particular to a protein anomaly detection method based on ultra-large-scale evolutionary algorithms and hardware acceleration. Background technique [0002] In recent years, the field of protein abnormality detection has attracted more and more attention. By detecting and classifying the abundance of a large number of proteins in human serum at a given mass value, it can be used in abnormal judgment assistance, drug treatment and other applications. [0003] Abnormal detection of proteins can help us understand the health of the human body, and with the continuous advancement of biotechnology, more and more features are collected and analyzed on proteins, so that a large number of features can be extracted. However, as the scale of data increases, the difficulty of selecting feature selection schemes for protein anomaly detection also in...

AI Technical Summary

Problems solved by technology

[0004] However, most of the current protein anomaly detection methods on large-scale mass spectrometry feature sets often have too much computation due to high-dimensional feature numbers, take a long time, and occupy a large amount of memory, so that they cannot efficiently obtain high-quality protein anomaly detection feature selection schemes; Although the existing method of grouping and reducing dimensionality of features can effectively reduce the amount of calculation and time consumption, but without much consideration of the interaction between features, it is easy to lose high-quality protein mass spectrometry features, thus generating inaccurate mass spectrometry feature selection scheme, leading to false detection of whether the protein is abnormal

Images