Integrated network analysis identifying potential novel drug candidates and targets for Parkinson's disease

Pusheng Quan,Kai Wang,Xinyu Zhang,Shi Yan,Chengqun Wei,Jingwei Cao,Lifen Yao,Shirong Wen

doi:10.1038/s41598-021-92701-2

Abstract

This study aimed to identify potential novel drug candidates and targets for Parkinson’s disease. First, 970 genes that have been reported to be related to PD were collected from five databases, and functional enrichment analysis of these genes was conducted to investigate their potential mechanisms. Then, we collected drugs and related targets from DrugBank, narrowed the list by proximity scores and Inverted Gene Set Enrichment analysis of drug targets, and identified potential drug candidates for PD treatment. Finally, we compared the expression distribution of the candidate drug-target genes between the PD group and the control group in the public dataset with the largest sample size (GSE99039) in Gene Expression Omnibus. Ten drugs with an FDR < 0.1 and their corresponding targets were identified. Some target genes of the ten drugs significantly overlapped with PD-related genes or already known therapeutic targets for PD. Nine differentially expressed drug-target genes with p < 0.05 were screened. This work will facilitate further research into the possible efficacy of new drugs for PD and will provide valuable clues for drug design.

Highlights

This study aimed to identify potential novel drug candidates and targets for Parkinson’s disease
We selected a systematic computation framework to explore potential treatment options for Parkinson’s disease (PD) based on existing data about diseases, drugs and drug targets
Since drugs usually interact with specific targets to exert an effect on biological processes, and drug targets always interact with disease-related genes, we collected PD-associated genes

Summary

Introduction

This study aimed to identify potential novel drug candidates and targets for Parkinson’s disease. New drug development is affected by many factors, and 85.1% of potential drugs for PD tested far have failed in the clinical trial p hase[6] Under such a situation, the repositioning of available drugs for other disorders as potential novel therapeutic agents for PD becomes an ideal approach. We present an integrated method for the comparisons of gene expression signatures between a disease model and drug-treated condition network, prediction of drugprotein interactions, and large transcriptomic dataset mining. The purpose of this scheme is to predict and identify new related drugs and targets by applying integrated network pharmacology and transcriptome analysis. Our work will facilitate further studies for better preventive strategies for PD

Objectives

Methods

Results

Conclusion