Sign In

Communications of the ACM

ACM TechNews

Researchers Use AI to Find Patterns Among Multitude of People, Cells


View as: Print Mobile App Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook
The proverbial needle in a haystack.

Yale University researchers have developed a way to leverage neural networks to reveal patterns of activity of individual cells from multiple individuals.

Credit: stock.adobe.com

Researchers at Yale University have developed a method of leveraging artificial intelligence (AI) neural networks to reveal larger patterns of activity of individual cells that come from several individuals.

The AI neural network, called SAUCIE (Sparse Autoencoder for Clustering, Imputation, and Embedding), can reveal minute cellular differences within individuals, as well as broader patterns that describe how the body functions.

The new method will allow researchers to identify larger clusters of cellular activity that could shed light on the basis of a host's pathogens.

For example, the team used SAUCIE to analyze 20 million cells from 60 patients and identify rare Gamma-Delta T cell types that regulate how the body responds to the virus that causes Dengue fever.

From YaleNews
View Full Article

 

Abstracts Copyright © 2019 SmithBucklin, Washington, DC, USA


 

No entries found