Mitigating the lack of knowledge about long noncoding RNA: extracting biological functions

Abstract

AbstractLong non-coding RNAs (lncRNAs) play crucial roles in diverse biological processes. For example, they help to regulate gene expression and to shape the 3D organization of the cellular nucleus. However, their functions in these processes are not well known. Most of the ongoing research efforts regarding lncRNAs have focused on predicting their properties and their functions. Here, we aimed to extract the biological functions of lncRNAs referred in the biomedical literature. To this end, we performed gene enrichment tests for 417 verified gene names extracted from the scientific literature. We also succeeded in extracting biologically significant functions from 2455 single sentences of 2513 scientific abstracts retrieved from the PubMed digital library. These sentences had to be filtered to eliminate those that did not mention a biological function. They were also split into 2-3-grams and the meaningful function annotations based on Gene Ontology terms were extracted. The results provided by the gene enrichment test and the knowledge gained by automatic extraction of information from the scientific literature showed that lncRNAs play a critical role in all biological and cellular processes. Although simple natural language processing techniques fulfilled our purposes of extracting concepts concerning the biological functions of lncRNAs, in the future, we aim to use more sophisticated methods, such as grammatical rules, to extract more functions and to generate a full database for the biological functions related to lncRNAs.