Drug relocation method based on differential expression data

Abstract

The invention provides a drug relocation method based on differential expression data, and mainly solves the problems of no screening of differential genes and low drug prediction accuracy in the prior art. The method comprises the following steps: acquiring gene data, calculating a gene expression change value and a significant value thereof, and screening out a gene set with significant differential expression; constructing two query gene sets of the disease by utilizing the gene set and the pathogenic gene; obtaining correlation values of the two gene sets and drugs and drug prediction accuracy corresponding to the correlation values; comparing the predicted drug accuracy of the two gene sets; calculating a difference value between the screened gene set and standard normal distribution,and constructing a prediction model of the optimal disease query gene set according to the difference value and the calculated expression threshold; using the model for predicting gene sets of otherdiseases and calculating the accuracy of predicted drugs, and screening out candidate drugs with potential treatment effects. The method is high in prediction accuracy and can be used for predicting candidate drugs with potential treatment effects on diseases.

Description

technical field [0001] The invention belongs to the technical field of data mining, and in particular relates to a drug repositioning method, which can be used to predict candidate drugs with good therapeutic effect. Background technique [0002] It is well known in the industry that the development of drugs takes a long time. It usually takes 11.4-13.5 years to develop a new drug from start to finish, and the cost of each drug is 16.1-180 billion US dollars. Despite the enormous time and money invested in drug development, more than 90% of drugs fail. For example, from 1949 to 2014, the US Food and Drug Administration (FDA) approved a total of 150 drugs for cancer treatment. Due to the existence of cancer subtypes and drug resistance, the available drugs are insufficient, so exploring new drug development strategies seems to be a problem. Extra important. Drug repurposing, as a strategy of repurposing old drugs, saves a lot of time and cost in drug development. Tradition...

AI Technical Summary

Problems solved by technology

Due to slight changes in molecular structure, its activity will be significantly improved or lost

Therefore, this type of method has a high false positive rate in the application

The second category is to use the spatial structure information of drugs to reposition drugs. This type of method mainly uses the spatial structure modeling of drug molecules and proteins to simulate their direct physical interaction, which relies on the analysis of drug molecules and protein structures. However, some of the currently resolved protein structures have errors and many important protein structures have not been completely decomposed, which makes the modeling of compound-protein interactions incomplete. Therefore, such methods also have high false positives

Moreover, these methods select genes with significant differential expression, especially those genes with high differential expression, as the query gene set of the disease, and do not consider the pathogenic genes that are very closely related to the disease at all, so that the accuracy of the prediction results is low

Images