Using MicroRNA Coding Sequences and Target Predictions to Improve Functional Prioritization of Non-coding SNVS
MicroRNAs (miRNAs) have been receiving growing interests over the past decade as an important class of regulatory RNA. MiRNAs have been shown to play a wide variety of functional roles in developmental and pathogenic processes. Like other genomic elements, miRNA functions are subject to the influence of single nucleotide variants (SNVs). With SNV data becoming more available, there is a growing need to understand the functional importance of miRNA related SNVs. To achieve this goal, there are two important and fundamental resources that should not be neglected: the miRNA coding sequences and miRNA target sites. First, the SNVs in miRNA coding sequences can impact hundreds of genes simultaneously. Second, SNVs located in the miRNA target sites can disrupt the regulatory effect of this miRNA, which have been shown to be related with various diseases. Thus, we explored the possible applications to better utilize these two types of miRNA information to help us understand genetic variants through three projects. ^ In our first project, we used a single-variant-based and gene-based approach to discover the association between SNVs located in miRNA-coding sequences or miRNA target sites and 17 cardiovascular disease risk factors. Incorporating these two components of additional information regarding miRNA, we found several significant associations, which were partially validated in an independent test set. ^ In our second project, we established a comprehensive database on SNVs located in putative miRNA target sites. By calculating three miRNA target related scores and retrieving multiple functional annotation scores, we curated the largest database on putative miRNA target site SNVs. This database can help studies easily and quickly identify putative SNVs that may impact miRNA targeting and facilitate the prioritization of functionally important SNVs in putative miRNA target sites at genome level. ^ In our third project, we utilized our database and an unsupervised learning approach to calculate an integrative score. Using a combined testing set, our calculated integrative score, MyEigen, outperformed all other scores in prioritizing functional SNVs located in putative miRNA target sites.^
Li, Chang, "Using MicroRNA Coding Sequences and Target Predictions to Improve Functional Prioritization of Non-coding SNVS" (2018). Texas Medical Center Dissertations (via ProQuest). AAI10928390.