Method for predicting cancer synthesis lethal gene pairs based on decision tree and linear regression model

Abstract

The invention relates to the field of gene prediction, and discloses a cancer synthesis lethal gene pair prediction method based on a decision tree model and a linear regression model. The method is mainly divided into a data training stage and a synthetic lethal gene pair testing stage. The method sequentially comprises the steps: firstly, extracting the data, which contains all mutant gene paircoverage rates, DNA methylation, mRNA expression profiles and copy number variation, from multi-omics data to serve as model feature values to be clustered, removing the false positive, performing normalization processing, and training a decision tree model and a linear regression model; secondly, respectively predicting synthetic lethal gene pairs possibly existing in various cancers through utilizing the decision tree model and the linear regression model to obtain distribution maps of the synthetic lethal gene pairs in different cancers; and finally, comparing the two models to obtain 508 pairs of synthetic lethal genes existing in the pan-cancer. According to the method, the synthetic lethal gene pairs possibly existing in various cancers can be accurately predicted, so a basis is provided for accurate treatment of the cancers.

Description

technical field [0001] The invention belongs to the field of gene prediction, and specifically relates to a combination of genes with synthetic lethal effects related to cancer, which is integrated and analyzed for multi-omics high-throughput sequencing data and predicted through different model algorithms based on various molecular level feature values. Background technique [0002] As we all know, the treatment of cancer has always been a difficult problem in the field of modern medicine. Traditional treatment methods are difficult to completely kill cancer cells due to the spread of cancer cells with the blood. In recent years, the development of genome sequencing technology has made some Emerging treatments continue to emerge and offer possible cures for cancer. Targeted therapy has attracted widespread attention due to its specific treatment methods and less damage to normal cells. However, the problem of drug resistance in targeted therapy widely exists in a variety of...

AI Technical Summary

Problems solved by technology

[0003] However, there are certain difficulties in the prediction of synthetic lethal gene pairs. So far, the results of various prediction models of synthetic lethal gene pairs that have been published are quite different. For example, the famous DAISY algorithm (Jerby-Arnon, L., Pfetzer, N., Waldman, Y.Y., Mcgarry, L., James, D., Shanks, E., Seashore-Ludlow, B., Weinstock, A., Geiger, T. and Clemons, P.A., 2014. Predicting Cancer-Specific Vulnerability via Data-Driven Detection of Synthetic Lethality.Cell158,1199-1209.), although this algorithm uses a variety of data screening criteria for screening, it still misses a large number of synthetic lethal gene pairs, and uses Fixed criteria for screening

In view of this feature, MiSL (Mining Synthetic Lethals) (Sinha, S., Thomas, D., Chan, S., Gao, Y., Brunen, D., Torabi, D., published by Subarna Sinha et al. in Nature Communications magazine .,Reinisch,A.,Hernandez,D

Images