Exploring Drugs For Treating Alzheimer’s Disease Using Computational Drug Repositioning

Alzheimer’s disease(AD)is a progressive neurological disease,which is the most common type of dementia;it destroys the patient’s memory and other important cognitive functions,causing the patient to experience continued mental decline and abnormal behavior.This severely affects the patients’ daily life and social functions,and at the same time brings heavy economic burden to their family and society.The pathological mechanism of AD is very complicated,often caused by a combination of multiple genetic factors,environmental factors and some psychosocial factors.To date,no drug has been effective in the treatment of AD.Taking into account the shortcomings of traditional drug like long development cycles,complicated process and huge cost,in this study,we apply a systematic computational drug repositioning approach to explore drug candidates that can be used for AD.Meanwhile,based on the current research results of genomics,proteomics,and bioinformatics,we hope to further understand the complex molecular mechanisms and their internal connections of AD.Method:(1)Collected risk genes of AD.The known AD-related genes were collected from OMIM(Online Mendelian Inheritance in Man),KEGG(Kyoto Encyclopedia of Genes and Genomes)and Phe Gen I(the Phenotype-Genotype Integrator).(2)Functional enrichment analysis.In order to reveal the complex molecular mechanism of AD,functional enrichment analysis was performed on its the risk gene set,including the terms of GO(Gene Ontology)biological processes and KEGG biological pathway entries;meanwhile,by analyzing the functional modules of the AD risk gene network,the dominant inheritance factors of AD were explored.(3)Constructed human proteinprotein interaction network and collected small drug molecules.The existing human protein-protein association information was collected from Hu RI(the Human Reference Protein Interactome),PINA(Protein Interaction Network Analysis platform)and STRING(the Search Tool for the Retrieval of Interacting Genes/Proteins)to construct a relatively complete human protein interaction network.In order to implement the drug relocation algorithm,detailed information of drugs was extracted from the Drug-Bank.(4)Calculated the distance between drugs’ targets and AD risk genes based on the PPIN(Protein-Protein Interaction Network).Calculating the correlation between drugs and diseases based on PPIN,so as to find potential drugs for diseases,was an effective strategy for drug relocation.In the study,we proposed a network algorithm based on the PPIN we had built,which could unbiasedly quantify the closeness of small drug molecules and AD gene set in the network.And screened potential AD treatment of single drugs by setting an appropriate threshold.(5)Inverse gene set enrichment analysis algorithm(IGSEA)to screen drug candidates for AD.Based on the hypothesis that the drug action was essentially reversing the pathological gene expression to return to normal,we proposed the IGSEA algorithm to further screen for drug candidates for treating AD after the network algorithm.(6)Evaluation.We evaluated whether the drug obtained by the above algorithm could be a candidate drug by reviewing the literature,learning about the main indications and prospective studies of the drug.Result:(1)The risk gene set of AD.A total of 561 AD risk genes were collected from the databases,including star genes such as APP,PSEN1,PSEN2 and APOE,as well as NDU families and UQCR families related to energy metabolism,and CHRNA families,HTR families related to synapse formation and transmitter transmission.(2)Functional enrichment analysis.Through the functional enrichment analysis,the biological processes and biological pathways involved in AD risk genes and their connections were identified.Among them,18 types of GO biological processes were enriched,including NADH dehydrogenase(ubiquinone)activity,amyloid β protein metabolism,and neurogenesis regulation;and 10 types of KEGG pathways,such as cholinergic synapses,dopaminergic synapses and numerous signaling pathways.(3)Candidate repositioning drug for AD.By collecting human protein interaction information from the databases,a network of 17,232 proteins/genes and 493,232 interaction relationships was constructed;5,595 drugs containing target information were extracted from Drug Bank.Through the network algorithm based on PPIN,1092 drugs with a close relationship with AD were screened out.By performing the IGSEA algorithm,24 drugs were finally retained as candidate repositioning drugs.(4)Drug evaluation.By consulting the literature and database,we found that indications of the 24 drugs including mental diseases,cardiovascular diseases,inflammation and cancer and other diseases and symptoms;among them,xaliproden was a drug being investigated for treatment in AD.Conclusion: The analysis of AD related risk genes through bioinformatics reveals the complex pathological mechanism of AD.In addition to the a priori process like amyloid metabolism control process,we also discover that energy metabolism processes,signal transduction pathways,especially the nervous system occurrence process has an important impact on the progress of AD.In short,the occurrence and development of AD involve major systems of the whole body and the combined effects of many biological processes.In addition,through the calculation of drug repositioning,we have screened dozens of possible drug candidates for the treatment of AD,and its indications are extensive.Of course,the medicinal value,usage and dosage of these drugs need further experimental proof.In addition,the algorithm used in the object is very effective for drug repositioning for complex diseases like AD,providing a new idea for calculating the direction of drug relocation.