Scientists and researchers have teamed up to test and evaluate the algorithms used for searching the cancer genome. Their work describes and classifies the strengths and weaknesses of more than 20 algorithms.
There is a new category of cancer-searching algorithms in town – these algorithms mine genetic information in cancer databases by focusing on internal gene structure. Because of this, they’re called Subgene resolution algorithms. Of course, there are the traditional approaches as well – but these algorithms focus on genes treated as single units (so they’re called whole-gene algorithms).
As Adam Godzik, Ph.D., director of the Bioinformatics and Structural Biology Program at SBP and the senior author of the study explains, despite increasingly popular high-res genome sequences, many consider a gene a single unit. The problem with this assumption is that there are various events that can occur within a gene, at subgene level. For this reason, subgene algorithms are usually a better option, as they provide a high-res view that can explain different mutations in the same gene.
The team applied subgene algorithms to the data from The Cancer Genome Atlas that includes genome data from 33 tumor types from more than 11,000 patients. They found that:
a) Subgene algorithms were able to reproduce the list of all known cancer genes which are established by cancer researchers;
b) Subgene algorithms were able to find some new cancer driver genes that were missed by whole-gene algorithms.
All in all, it appears that Subgene algorithms are the future when it comes to searching for cancer mutations.
Source:
ScienceDaily (https://www.sciencedaily.com/releases/2017/07/170717125715.htm)
The Wonderful World of Algorithms 