Innovative research, published in PLOS Computational Biology this week, explains how thousands of "previously ignored genetic mutations" may contribute to the growth of malignant tumors. Using a new statistical approach, scientists find new patterns in proteins.
Modern pharmaceutical interventions have been designed to exploit our knowledge of certain cancer-related mutations; they target proteins that are altered due to the mutations in the genes that code them.
To date, only a tiny number of these important cancer-causing mutations have been pinpointed.
Although studies using large numbers of participants have identified statistically significant mutations, even somatic variants that are considered to be important drivers of cancer appear in relatively low frequencies.
Similarly, many more mutations have been noted that do not quite reach a reliable enough level of statistical significance.
A group of researchers from the University of Maryland in College Park, led by Thomas Peterson, used a new method of statistical analysis to tackle this gap in our knowledge and examine similar mutations that are spread across the genome.
Using genetic data, they targeted similar mutations "shared by families of related proteins," specifically examining mutations in subcomponents of proteins known as protein domains.
Perusing protein domains
Protein domains are distinct units within proteins - each domain carries out specific functions independently from the other domains that make up the same protein. Proteins, even if they are coded by different genes and carry out completely different physiological roles, can share common protein domains.
The research team utilized existing knowledge about protein domain function and structure to identify regions within the domains where mutations might be more likely to occur in tumors. Their work makes use of the great swathes of data produced by previous gene research. As the authors of the study write, "we leverage decades of important findings from structural genomics."
In other words, rather than focus on mutations in single regions of specific genes, the team focused on mutations that occurred in similar regions across families of proteins.
In all, they collected data about somatic variants from 5,848 patients in The Cancer Genome Atlas, which included patients with 20 different cancer types.
Using this fresh approach, the team unearthed thousands of rare tumor mutations that occur in the "same domain location as mutations found in other proteins in other tumors." This implies that they are involved in cancer.
"Maybe only two patients have a mutation in a particular protein, but when you realize it is in exactly the same position within the domain as mutations in other proteins in cancer patients, you realize it's important to investigate those two mutations."
Maricel Kann, senior author
The authors refer to these protein domains that are likely to comprise cancer-causing mutations as "oncodomains." Understanding more about oncodomains could eventually lead to improved cancer treatments. As Kann explains: "Because the domains are the same across so many proteins, it is possible that a single treatment could tackle cancers caused by a broad spectrum of mutated proteins."
Although this work represents the first step in a new direction, it is a significant first step that promises to improve cancer treatment in the long run. Investigating the genetic basis of cancer from a different standpoint gives other researchers a new angle from which to approach the problem.
The authors conclude: "Determining which variants are most important for tumorigenesis will help elucidate the mechanisms driving tumor progression and could ultimately provide a new set of drug targets for families of genes that display similar variation at the structural and functional level."