Exclusive
scalehealthtech Realize your Healthcare’s Digital Transformation journey with ScaleHealthTech Learn More

MIT Scientists Develop AI-Powered Model to Predict 3D Genome Structures

Written by : Jayati Dubey

February 5, 2025

Category Img

MIT’s ChromoGen model leverages advances in deep learning and generative AI to rapidly predict chromatin conformations.

A team of researchers at the Massachusetts Institute of Technology (MIT) has developed an artificial intelligence model capable of predicting the three-dimensional structure of the genome with unprecedented speed and accuracy.

This innovative approach, detailed in a study published in Science Advances, offers a faster and more efficient alternative to existing experimental methods, potentially transforming genetic research and precision medicine.

The AI model, known as ChromoGen, can generate thousands of chromatin structure predictions within minutes, whereas traditional methods take weeks to analyze a single cell.

This will allow scientists to better understand how genome organization influences gene expression, paving the way for new insights into cellular functions, genetic disorders, and disease mechanisms.

The Role of AI in Predicting Genome Folding

MIT’s ChromoGen model leverages advances in deep learning and generative AI to rapidly predict chromatin conformations.

By analyzing DNA sequences and chromatin accessibility data, the AI can generate physically accurate 3D genome structures that closely resemble those obtained through experimental techniques.

"Our goal was to try to predict the three-dimensional genome structure from the underlying DNA sequence," says Bin Zhang, associate professor of chemistry at MIT and senior author of the study.

"Now that we can do that, which puts this technique on par with the cutting-edge experimental techniques, it can really open up a lot of interesting opportunities."

Graduate students Greg Schuette and Zhuohan Lao, the study’s lead authors, highlight that deep learning excels at pattern recognition, allowing the AI to detect key structural information embedded within long DNA sequences.

The model effectively predicts how chromatin folds within different cell types, offering insights into gene regulation, disease development, and potential therapeutic targets.

Overcoming Limitations of Traditional Techniques

Historically, genome structures have been studied using Hi-C sequencing, an experimental method that maps physical interactions between DNA segments.

While Hi-C provides valuable insights, it has several drawbacks, including high costs, labor-intensive protocols, and long processing times. Moreover, Hi-C only offers a snapshot of chromatin organization, failing to capture the full range of structural variations present in living cells.

ChromoGen circumvents these limitations by offering fast, AI-driven predictions.

Using generative AI, the model analyzes large DNA sequences to infer chromatin organization, generates thousands of possible conformations for each region of the genome, and captures dynamic genome folding patterns, providing a more comprehensive view of chromatin structure.

"Whereas you might spend six months running experiments to get a few dozen structures in a given cell type, you can generate a thousand structures in a particular region with our model in 20 minutes on just one GPU," says Schuette.

Potential Applications in Medicine & Genomics

The ability to rapidly predict chromatin conformations has far-reaching implications for medical research and biotechnology. One key application is in studying cell-specific gene regulation.

By analyzing variations in chromatin structures across different cell types, researchers can gain deeper insights into gene expression control mechanisms, enhancing the understanding of cell differentiation, tissue development, and stem cell biology.

Another significant area of impact is investigating genetic disorders and cancer. Many diseases, including cancers and neurological conditions, are associated with abnormal chromatin folding.

ChromoGen could help identify structural changes linked to these diseases, paving the way for improved diagnostic tools and targeted therapies.

Assessing the impact of DNA mutations is also a crucial application. By modeling how genetic mutations alter chromatin conformation, researchers can explore disease mechanisms and potential drug targets. This is particularly valuable for advancing personalized medicine and gene therapy approaches.

Additionally, understanding chromatin dynamics plays a vital role in drug discovery and gene editing.

Predicting how DNA modifications influence genome organization can refine gene-editing technologies such as CRISPR, enhancing the precision and effectiveness of genetic therapies.

The MIT team has made ChromoGen’s data and AI model publicly available, allowing researchers worldwide to explore its capabilities. The study, funded by the National Institutes of Health, represents a major step forward in AI-driven genomics.

Stay tuned for more such updates on Digital Health News.


POPULAR CATEGORIES

WEEKLY POPULAR POSTS

ABOUT US

Digital Health News ( DHN) is India’s first dedicated digital health news platform launched by Industry recognized HealthTech Leaders. DHN Is Industry’s Leading Source Of HealthTech Business, Insights, Trends And Policy News.

DHN Provides In-Depth Data Analysis And Covers Most Impactful News As They Happen Across Entire Ecosystem Including Emerging Technology Trends And Innovations, Digital Health Startups, Hospitals, Health Insurance, Govt. Agencies & Policies, Pharmaceuticals And Biotech.

CONTACT US

© Digital Health News 2025