PriorGene Suite: A Tool For Candidate Gene Prioritization And Pathway Enrichment Analysis For Human Complex Diseases

Objective:With the development of genomics and related high-throughput technologies,we are increasingly capable of exploring mechanisms of human complex diseases and related biological processes at the molecular level.Corresponding to this,we have also accumulated a large number of genes and its function information associated with various diseases.However,for a specific disease,genes identified by different technologies and methods,most of which may play a minor or indirect role in the pathogenesis or development of the disease,only a few genes are closely related to it.Therefore,how to integrate the relevant candidate gene information accumulated by different platforms and technologies to screen out the most relevant gene sets,and explore its function to reveal the molecular mechanism of disease development at the systematic level,is still a challenge in the study of complex diseases.In recent years,although many methods have been developed to solve these problems,they all have certain defects or limitations.Therefore,according to the characteristics of complex diseases,it is of great significance to develop more effective candidate gene prioritization methods and corresponding gene function analysis tools.Method:1.PriorGene:Candidate gene prioritization.We collected and integrated disease-associated candidate genes from various study categories.And then used simulated annealing to assign optimal weight to each study category to measure its contribution to the genes correlation with the phenotype.According to the combined score of the candidate genes and the threshold we set,a set of optimal disease-related gene sets were finally screened out.At the same time,we performed functional enrichment analysis of the selected genes and obtained their significantly enriched biological pathways.Furthermore,the prioritized results of PriorGene,GLAD4U,Phenolyzer,and GeneCards were compared and analyzed using the Receiver Operating Characteristic Curve?ROC?.2.PriorFun:Pathway enrichment analysis.In this study,we considered the constitutive non-equivalence of different genes in real biological pathways.We utilized the functional interactive information to weight genes in the pathway based on the assumption that if a gene is associated with more genes than expected within the pathway,then it is more likely to be functionally important in this pathway.This correlation feature was estimated using a hypergeometric sampling model.Then we integrated the gene weights into the right-tailed Fisher exact test to calculate the significance of different biological pathways.Results:1.Candidate gene prioritization.We obtained 278,2158,694 and 1635schizophrenia-related candidate genes from four research sources,i.e.,association studies,linkage analysis,gene expression analysis and literature search,respectively.A total of 267 genes?referred to as SCZgenes?were prioritized by using the combined score of 0.653.Functional enrichment analysis found that many of the biological processes involved in neurodevelopment,synaptic transmission,and neural signaling were significantly enriched in SCZgenes,such as synaptic transmission,regulation of transmission of nerve impulse,glutamate signaling pathway,cell surface receptor linked signal transduction and G-protein coupled receptor protein signaling pathway.There were also some biological processes related to drug reactions,cognition and memory.The result of pathway enrichment analysis showed that most of the significantly enriched pathways are related to neurological function or drug addiction,such as dopaminergic synapses,glutamatergic synapses,serotonergic synapses,cocaine addiction,amphetamine addiction,nicotine addiction and alcohol addiction.And ROC curve of different tools showed that AUC_PriorGene?0.939?>AUC_GeneCards?0.917?>AUC_Phenolyzer?0.903?>AUC_GLAD4U?0.793?.2.Pathway enrichment analysis.A total of 177 nicotine addiction-related genes?NAgenes?and 34 smoking initiation-related genes?SIgenes?were obtained by screening out the reports related to smoking behavior in PUBMED.Pathway enrichment analysis of NAgenes and SIgenes using ORA and network-based gene weighting over-representation analysis showed that most of the significantly enriched pathways were associated with the nervous system,which was consistent with previous findings.And network-based gene weighting over-representation analysis could detect more pathways that were truly related to the disease under study.Conclusions:1.The results of the functional and pathway enrichment analysis are in good agreement with the existing studies,demonstrating that the schizophrenia-related genes prioritized by PriorGene are credible and can serve as candidate genes for further study of the molecular mechanisms underlying schizophrenia.2.PriorGene were compared with GLAD4U,Phenolyzer and GeneCards by using ROC curve,as a result,it had best performance according to the AUC of each curve.3.Schizophrenia is a complex disease affecting the function of the nervous system.Drug addiction may have a causal relationship with schizophrenia,and schizophrenia may directly or indirectly affect advanced neurobiology processes such as learning,cognition and memory.4.In this study,the obtained NAgenes and SIgenes were pretty reliable,which provided a valuable gene set for the future experimental research and corresponding database construction of nicotine addiction.