A new machine learning algorithm can predict how genes are regulated in individual cells, a breakthrough that could lead to new treatments for a variety of diseases.

The algorithm, developed by researchers at the University of California, Berkeley, is able to identify the specific DNA sequences that control the expression of genes. This information could be used to develop drugs that target these sequences and either turn genes on or off.

"This is a major breakthrough in our understanding of how genes are regulated," said study leader John L. Rinn, PhD, an associate professor of molecular and cell biology at UC Berkeley. "This new algorithm will allow us to identify the key regulatory elements in the genome and develop new therapies for a variety of diseases."

The algorithm, called cis-BPNet, was trained on a large dataset of gene expression data from different cell types. The algorithm was able to learn the relationships between the DNA sequences and the expression of genes, and it can now predict how genes will be expressed in different cell types.

The researchers tested the algorithm on a variety of genes, and they found that it was able to accurately predict the expression of genes in different cell types. The algorithm was also able to identify the key regulatory elements in the genome that control the expression of genes.

This information could be used to develop drugs that target these regulatory elements and either turn genes on or off. This could lead to new treatments for a variety of diseases, such as cancer, diabetes, and heart disease.

"This is a powerful new tool that will allow us to understand how genes are regulated and develop new therapies for a variety of diseases," said Rinn.

The study was published in the journal Cell.