Massachusetts Institute of Technology researchers present a deep convolutional neural network model designed to discern the DNA methylation consequences of non-coding DNA variants based on learned regulatory codes and neighboring sequence features. The approach, dubbed CpGenie, "predicts the impact of sequence variants on DNA methylation with an accuracy that surpasses existing methods for functional variants prioritization," the study's authors say.

Get the full story

This story is free
for registered users

Registering provides access to this and other free content.

Register now.

Already have an account?
Login Now.

While gene therapies may have high price tags, they could be cheaper than the cost of managing disease, according to MIT's Technology Review.

Researchers are looking for markers that indicate which cancer patients may respond to immunotherapies, the Associated Press writes.

In Nature this week: paternal age associated with de novo mutations in children, and more.

Nature News writes that researchers are still wrangling over the role of the p-value.