Clustering method of gene microarray containing missing values based on joint training

Abstract

The invention discloses a clustering method for gene microarrays containing missing values based on joint training; the method comprises the following steps: calculating the missing rate of gene microarray data, removing gene points with the missing rate exceeding 10%, and then dividing the gene microarray data into a training set and a test set; constructing a deep neural network, including a Sequence-To-Sequence encoding-decoding network and an adversarial learning module, and training the constructed deep neural network by using the training set to determine parameters of the deep neural network; inputting the test set into a deep neural network to obtain deep feature representation containing missing value gene microarray data in the test set; and finally, applying a K-means clustering algorithm to the deep feature representation to obtain a clustering result of the gene microarray containing the missing values. According to the method, an end-to-end framework can be provided for gene expression data clustering, and the difficulty that a proper filling method and a clustering method need to be selected to be combined in a traditional method is solved.

Description

technical field [0001] The invention relates to the technical field of gene microarrays, in particular to a clustering method for gene microarrays containing missing values ​​based on joint training. Background technique [0002] Clustering methods have always been of great experimental value in gene microarray research. For example, after obtaining the expression sequences of several genes, clustering the expression sequences of genes is a common experimental method. function to help. [0003] In the process of obtaining gene expression sequence data, missing values ​​often appear due to human omissions or uncontrollable experimental errors. The appearance of missing values ​​has greatly hindered the downstream research of gene microarrays. Resetting the experimental environment to obtain a complete gene expression sequence is expensive and impractical. Therefore, how to mine information that is beneficial to analysis from gene expression data with missing values ​​is v...

AI Technical Summary

Problems solved by technology

In addition, for clustering tasks, how to select the best combination of imputation method and clustering method for gene microarrays with missing values ​​in an unsupervised environment is also a difficult problem

Images