Computational Biology and Genomics

The sequencing of the human genome is considered a monumental achievement in genetics and has unveiled the blueprint of our genetic makeup. The discipline of computational biology uses data analysis, modeling and simulation allowing the fast changing field of genomics to produce new insights using genomic sequencing data.

One particular area especially important for diagnosis and treatment of certain diseases is, gene encoding. Genes have a well-defined structure made up of exons, which represent the coding part, and introns, which generally represent the non-coding part. Most exons of a gene are coded for protein, however, there is a small number of them that may still be non-coding, especially the ones either at the beginning or at the end of the gene. While a lot of progress has been made in understanding the genes and the coded parts, there is less known about these noncoding genes in our genome and how that contributes to human health. Non-coding genes, specifically non-coding exons, produce RNA that is not translated to proteins and is typically dismissed as “junk DNA.” However, as we start to unpack more about RNA therapeutics, research has started to recognize the hidden potential in noncoding RNA and identified that as a major target point for further study. Some references here and here

To study the impact of genetic variants of non-coding genes on human health, we need to start by ranking and prioritizing them since the interpretation of non-coding genes has been fairly difficult. 

Since the past few years, the Kellis Lab at MIT Computer Science & Broad Institute has focused on using mathematical and data analysis models to look at the probabilities for changes in protein expression and predict any premature exon endings for protein-coding exons. Much of the Kellis Lab has focused on the importance of these proteins and interactions in building therapeutics and more targeted ones. However, lately there is also ongoing research in understanding more about non-coding exons and their impact on human disease and health.  

I was fortunate to spend some time at the lab last semester understanding the role of non-coding RNA made of non-coding exons, and comparing results between protein-coding, non-coding, and intergenic SNP to recognize the differences and similarities. This was largely conducted through mathematical models, written in R, to evaluate and build models to predict RNA structures that impact human health considering factors such as RNA folding patterns, interactions with proteins, and their involvement in regulatory networks.

The exploration of non-coding RNA within the framework of the Kellis Lab's research represents a significant stride toward unraveling the complexities of our genetic architecture. As we improve our mathematical models and compare tiny variations in our DNA across different areas, we uncover the subtle roles these mysterious sequences play in our health and diseases. 

What does this mean for the healthcare industry? 

The future of genomics combined with power of computation looks promising with the idea of using RNA for diagnostics and treatments, bringing new ideas to what we consider “personalized treatment”. With evolving progress in fields of AI and analytics, can imagine a new era where so-called "junk DNA" is seen as a valuable collection of biological regulators and potential drug targets. 

Computational biology can offer interesting opportunities for anybody looking to tackle the discovery of the next generation of disease diagnosis and treatments.

 We are just beginning an exciting journey towards a future of personalized medicine that can help many, and there is much more to come.