Machine Learning Program Offers Insights Into Promoter Sequences and Transcription

A team led by researchers at UT Southwestern Medical Center developed deep learning models to identify a simple set of rules that govern the activity of promoters – regions of DNA that initiate the process by which genes produce proteins. Their findings, published in Science, could lead to a better understanding of how promoters contribute to gene regulation in health and disease.

Although promoters are essential for the function of every gene, our understanding of how these genetic elements operate is incomplete despite decades of study that have defined many of their features. Our research sheds new light on how these sequences work in humans and other mammals."

Jian Zhou, Ph.D., Assistant Professor in the Lyda Hill Department of Bioinformatics at UT Southwestern

Dr. Zhou co-led the study with first author Kseniia Dudnyk, a graduate student in the Zhou Lab, and Jian Xu, Ph.D., a former researcher at the Children's Medical Center Research Institute at UT Southwestern.

Creating the proteins that cells use to perform their activities starts with a process known as transcription. That's when an RNA polymerase protein latches onto a DNA strand and copies – or transcribes – the encoded information into an RNA molecule. The region where the RNA polymerase attaches to begin transcription is called the promoter. In humans, promoters are typically composed of hundreds of base pairs, the units that make up DNA. Although researchers have identified common base pair sequences shared among some regions of DNA that are promoters, these sequences are often absent in human promoters, leaving the rules of how DNA sequence directs the transcription process unclear.

To better define promoters in humans and how they operate, the researchers developed a machine learning program they named Puffin. After analyzing data from tens of thousands of recognized human promoters, the program determined that they are made of three types of sequence patterns: motifs, initiators, and trinucleotides.

Puffin showed that depending on how these elements are arranged, they can activate or repress transcription of a gene. Puffin also can predict how the arrangement of these elements can direct RNA polymerase to preferentially transcribe a single strand of DNA or transcribe both strands simultaneously toward opposite directions. This bidirectional transcription is common in human genes.

The program further showed that mice and other mammals shared similar rule sets for governing promoter operation. In addition, Puffin allowed the researchers to predict whether and how transcription would occur if they mutated promoters, findings that closely matched those from experiments.

The study authors suggested that Puffin could help them understand how promoters work in healthy cells as well as how disease-associated alterations in promoters could lead to changes in gene transcription. This program is available on a free web server ( so that other researchers can test any promoter sequence of interest. They added that using a similar machine learning approach could offer insights into other facets of the genome that are still not well understood.

Additional UTSW researchers who contributed to this study are Donghong Cai, Ph.D., a joint postdoctoral fellow with the Xu Lab, and Chenlai Shi, M.S., Data Scientist.

This study was funded by grants from the National Institutes of Health (DP2GM146336, R01DK111430, R01CA230631, and R01CA259581); the Cancer Prevention and Research Institute of Texas (RR190071); and a Leukemia & Lymphoma Society Scholar award.

Dr. Zhou is a Lupe Murchison Foundation Scholar in Medical Research and a member of the Harold C. Simmons Comprehensive Cancer Center.

Journal reference:

Dudnyk, K., et al. (2024). Sequence basis of transcription initiation in the human genome. Science.


The opinions expressed here are the views of the writer and do not necessarily reflect the views and opinions of AZoLifeSciences.
Post a new comment

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.

You might also like...
New Gene Drive System Offers Hope for Sustainable Agriculture