RNA-seq Co-Expression for Functional Prediction of Non-Coding Genes скачать в хорошем качестве

RNA-seq Co-Expression for Functional Prediction of Non-Coding Genes

Noncoding (nc)RNAs are increasingly gaining appreciation for playing important roles in normal physiology and disease. However, there is still limited knowledge about their biological roles. Insights into the biological roles of ncRNA can be gained through analysis of gene expression correlation with the underlying assumption that highly correlated genes share similar functions. Here we expanded upon our prior efforts to predict gene function via RNA-seq gene expression correlations to predict the function of many ncRNAs. Utilizing pairwise correlations of coding and noncoding genes from The Cancer Genome Atlas (TCGA), which contains reads mapped to ncRNAs, and reference gene set libraries such as KEGG, GO and MGI, we performed gene function predictions for both coding and noncoding RNAs. The prediction for both coding and noncoding RNAs are served as a web-based application via an appyter. The appyter provides users with the ability to enter a human gene symbol, and in return, the user receives a custom Jupyter notebook report centered on their chosen gene. The report includes sorted predicted functions, top 100 most correlated genes, and a link to analyze these gene sets in Enrichr. For coding genes, the report compares predictions made via the TCGA data to predictions made via ARCHS4. Receiver operating characteristic curves are displayed side-by-side. Using the appyter, biomedical researchers can explore possible functions of ncRNAs of interest and generate hypotheses for further investigation. This presentation is by Christine Yoon, an undergraduate student at Duke. Christine describes her summer research project with the BD2K-LINCS DCIC in the Ma'ayan Lab at the Icahn School of Medicine at Mount Sinai. https://appyters.maayanlab.cloud/Gene... https://appyters.maayanlab.cloud https://icahn.mssm.edu/labs/maayan