Ideas Lab: Collaborative Research: Integrating cross-kingdom lncRNA genetic and functional interactions to build predictive network models

点击翻译按钮获取中文摘要

Although most well-known genes in cells include instructions for creating proteins as a functional product via an RNA intermediate, scientists are discovering an increasing number of genes that create non-protein coding RNA that are essential for the health and life cycle of organisms. Long non-coding RNAs (lncRNAs) are relatively recently recognized as important, and the exact role of most remains poorly understood. Detailed study of each lncRNA would provide essential information about possible functions, but would be time, labor, and resource intensive. Computational analyses can provide an efficient way to analyze many different molecules simultaneously, classify the molecules based on different shared features, and use these classifications and other information to predict function. This project will develop and use computational tools to integrate many different data types generated by previous studies and predict functions of lncRNAs in plant and animal cells. These predictions will allow for hypotheses about how the loss or increase of specific lncRNAs might impact a cell or whole organism. These hypotheses will then be tested through experimentation as a way to assess the efficacy of the predictions. Project resources will also support public education activities, K-12 teacher training on these emerging scientific topics, and student participation in modern computer and molecular genomics research.A central hypothesis of this project is that a multimodal, causal network analysis approach, constrained by physiological and molecular data from data-rich experiments, can be utilized to functionally annotate lncRNAs and to build a predictive model for future experiments. The ultimate goal of this project is to develop network-based models that can predict lncRNA function based on genetic, molecular, and genomic interactions, which can then be applied to datasets from any species. Through this award, researchers will construct and test networks that depict and predict lncRNA functions. Publicly available high-throughput sequencing from human cell lines, associated -omics data, and literature-based functional information will be used to infer a fully directional gene-regulatory network (GRN). Machine learning models will be applied to predict essential lncRNAs based on network topologies and connectivity. A similar approach will be applied to available datasets in the model plant, Arabidopsis thaliana. Predictions from both networks will be tested by manipulating ncRNA expression and assaying the impact on the network using Nanostring technology.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.