Genomic Variant-pathway Annotation Method And Its Applications In Autoimmune Diseases

Functional annotation methods of variants were employed in revealing the etiology of complex diseases.In recent years,genome-wide association study(GWAS)has been performed to identify millions of diseases associated variants.Linking variants to its related target genes and pathways has brought us closer to the goal of understanding complex diseases,but annotation of the majority of non-coding variants still present a particular challenge.Enrichment analysis widely used in interpret lager molecular datasets may be suboptimal for exploring underlying biological processes for gene interaction networks information wasn’t taken into consideration for those methods.Therefore,A systemic annotation method to explore functions of variants is of significance.Autoimmune Diseases(ADs)are a family of complex disorders with high heterogeneity,which contains at least 80 types of chronic diseases.Although the explicit cause remains unknown,investigations have revealed similar underlying mechanisms characterized by immune responses dysfunction.Genetic variation in autoimmune patients contributes to disease risk,and disease differentiation is closely related to genetic diversity.Mixed Connective Tissue Diseases(MCTD)can develop overlap autoimmune symptoms along with differentiation,which combines with signs from Systemic Lupus Erythematosus,Systemic Sclerosis and Myositis.Revealing associate pathogenesis requires understanding of complex biological processes,which would contribute to accurate diagnosis of diseases.Here we present a variants annotation pipeline comprised of two main modules: variationgene and gene-pathway projection.For variant-gene module,linkage disequilibrium were first taken into consideration to expand diseases related variants,which can be neglected in regression computational processes.Then variants located within a certain range of coding genes were allocated to the nearest gene.The target gene sets were computed with gene expression data in regression model to validate associated tissues.Moreover,e QTL and Hi C annotation projected variants to effect differentially expressed genes and chromatin interacting target genes respectively.Besides,the supplementary annotation of TFBs,conservative scores and chromatin states,and utilization of popular variant pathogenicity estimation tools help reduce the false discovery ratios.Gene-pathway projection integrates traditional enrichment method and entropy-based method to identify most significant disease specific pathways and gene sets.Entropy-based method that utilizes protein-protein interaction information can obtain pathways with strong interacting subnetworks related to diseases.The practice of pipeline in autoimmune diseases has proven more target genes could be identified for non-coding variants,and entropy-based pathway annotation methods distinguished most significant biological processes.For instance,BTN genes family enriched in MCTD-related diseases play a necessary role in γδT cells’ small phosphorylated nonpeptide antigens recognition.PHF8 target down-regulated genes and genes with H3K4 me and intermediate-Cp G-density promoters showed high entropy-based method significance.Histone demethylase PHF8 and H3K4me3 work as coactivator to regulate gene transcription,the abnormal interaction of which can be underlying mechanism for immune system disorders.Transcription factor Foxp3 target genes restricted expressed in CD4+ T cells controls cell lineage development,function loss of Foxp3 can result in dysfunction of immune tolerance.For the pipeline involves multiple tools and complicated data process programs,we present a python-based tools for variants functional annotation,which combines API assembly and local algorithm development.With small size of dependency datasets,the program is portable for multiple platforms.To summarize,further investigation of functional annotation pipeline demonstrated higher accuracy and applications in autoimmune diseases revealed diseases-associated biological processes.The disease-specific genes and pathways could be valuable for diseases classification and prediction.