Computational prediction of protein-coding regions in human transcriptomes: An application to the elderly

摘要

In RiboNucleic Acid (RNA) biology, it's been understood that cataloging the diversity of RNA transcript species in human cells is important before we can know the functions that these RNA molecules perform in the human body. While some RNA transcripts are protein-coding, others are non-coding transcripts. All of these RNA transcripts perform some functions in the human cells and therefore contribute to a number of biological processes in the body. Discriminating protein-coding RNA is fundamental to determining the genes encoded by these transcripts but existing computational tools that solve this task do not generalize as much. We therefore use a computational approach to more accurately predict protein-coding regions in these transcripts and with a good generalization performance. In this work, we propose a new application for our prediction technique that allows an alignment-independent discrimination of the protein-coding regions of RNA transcripts for an elderly population. Following an accurate computational discrimination of these transcriptome regions, we can then perform experiments in the laboratory in order to precisely find out which regions encode specific genes responsible for phenotypes like ageing in elderly people. Insights obtained from this work have the potential of opening the way to identification of novel biological roles for RNA-Seq data thereby making this computational approach to lead to an improvement in transcriptome analysis.