Communications of the ACM
Model Helps Identify Mutations That Drive Cancer
Currently, at least 30% of cancer patients have no detectable driver mutation that can be used to guide treatment.
Credit: Dylan Burnette and Jennifer Lippincott-Schwartz/National Institutes of Health
A team of researchers led by the Massachusetts Institute of Technology (MIT) has created a probabilistic computer model that rapidly scans the genome of cancer cells to identify cancer-driving mutations.
MIT's Maxwell Sherman said the deep neural network can accurately model the number of passenger mutations, enabling researchers to analyze the genome for regions with an unexpected accrual of possible driver mutations.
The researchers fed the model genomic data from 37 types of cancer, so it could measure the background mutation rates for each type.
The researchers discovered additional mutations that appear to play a role in tumor growth in 5% to 10% of patients, which could help clinicians to identify medications that would have better odds of treating those patients.
From MIT News
View Full Article
Abstracts Copyright © 2022 SmithBucklin, Washington, DC, USA
No entries found