Omics-mining Based Tumor Driver Mutation Prediction And Rare Disease Drug Repositioning

The rapid development of high-throughput omics sequencing technology provides essential technical support for our complex and rare diseases research.Developing effective data mining algorithms for large-scale omics data for drug discovery target and subsequent drug intervention for complex and rare diseases has become the core of precision medicine resarch.In our research,omics data mining algorithms for genomic,transcriptomic,GWAS omics data have been developed for drug target discovery and drug reposition for complex and rare diseases.The main content of our research is shown below.1.Research on driver gene prediction for complex diseases based on omics mining.Cancer is considered as the prototype study here.The mechanism and treatment of tumor are the most challenging problems in clinical.The prediction of tumor driver genes is the key part in tumor mechanism,drug target discovery,drug treatment and drug intervention research.Owing to the development of high-throughput sequencing technology in recent years,it has become possible to use patient exome sequencing data to identify driver genes in somatic mutations.However,the false positive rate of prediction results of these tools are still high.Based on tumor genomic data,our research developed an end-to-end,accurate and efficient driver gene identification tool for somatic mutation-derived driver genes:C~3(Consensus Cancer Driver Gene Caller).C~3provides a unified data input format,flexible and customizable parameters,and integrates six cancer driver gene identification tools including four categories of methods.The automated and integrated analysis of C~3can improve the accuracy and performance of the final prediction results.C~3is freely accessed at http://drivergene.rwebox.com/c3.2.Research on drug target discovery and drug repositioning based on omics mining of rare disease.With the development of next-generation sequencing,genomics,transcriptomics,and epigenetics data for common human diseases has accumulated rapidly.The establishment of public clinical samples databases for diseases have provided a large amount of data and user-friendly analysis platform for the study of pathogenic mechanism,treatment plans and survival prognosis.However,the data of rare diseases and some metal diseases is limited due to data security and privacy concerns,which hinders the development of orphan drugs for rare diseases.The establishment of the GWAS Catalog database makes up for the lack of rare disease data.In this paper,9 kinds of rare disease data were screened from the GWAS Catalog database guided by the"First Batch of Rare Disease Catalogue"issued by the National Health Commission of the People’s Republic of China.Analysis of drug repositioning for 9 rare diseases through the Meta Xcan and 7 similarity algorithms.The clinical phase III and clinical phase IV of related rare diseases have been collected as the gold standard dataset for evaluating drug repositioning results.Seven drug repositioning algorithms based on similarity alignment were systematically evaluated and tested.Collectively,our research can predict drugs with different mechanisms from known treatments without assuming any known drug targets or drug-disease relationships,which provides an effective computational framework for drug repositioning for rare diseases.In summary,based on omics data mining,our research conducts computational methodological research and exploration in disease target discovery and drug intervention for various diseases.Our research provides methodological support and guidance for precision medicine researchs on complex and rare diseases.