New AI System Uncovers Hidden Cell Subtypes to Advance Cancer Immunotherapy

These differences significantly impact how tumors respond to treatments, making accurate identification crucial. Traditionally, scientists have examined the RNA or protein molecules each cancer cell expresses, where it is located in the tumor, and what it looks like under a microscope. This has left a gap in understanding how cells behave in relation to their surroundings, limiting the precision of cancer treatments. The problem is compounded by methodologies that miss critical molecular or contextual information, such as identifying immune cells at the boundaries of tumors. Now, a new solution integrates multiple aspects of cancer cell biology, using deep learning technology to provide a comprehensive profile of individual cells, even those that seem similar but behave differently depending on their environment.

The solution, called CellLENS, was created by researchers from MIT (Cambridge, MA, USA) in collaboration with several other prestigious institutions. It employs a combination of convolutional neural networks and graph neural networks to integrate information on cell morphology, location, and behavior. By analyzing cancer samples, including those from lymphoma and liver cancer, the system identifies rare immune cell subtypes and their role in disease processes. This method enables scientists to group cells based on their biology and better understand their functions. The tool can detect important layers of information such as where a cell is located in tissue and how it interacts with its surroundings, helping to uncover previously overlooked cell behaviors.

The researchers tested and validated CellLENS by applying it to samples from both healthy tissue and various cancers. Their findings, published in Nature Immunology, revealed insights into the immune system's interaction with tumors, including rare immune cell subtypes and their role in tumor infiltration and immune suppression. These discoveries hold the potential to guide more targeted cancer treatments and improve immunotherapies by providing deeper insights into cellular behavior and location. Moving forward, the researchers aim to refine and expand the tool’s applications to accelerate the development of personalized therapies.

“I’m extremely excited by the potential of new AI tools, like CellLENS, to help us more holistically understand aberrant cellular behaviors within tissues,” said co-author Alex K. Shalek. “We can now measure a tremendous amount of information about individual cells and their tissue contexts with cutting-edge, multi-omic assays. Effectively leveraging that data to nominate new therapeutic leads is a critical step in developing improved interventions. When coupled with the right input data and careful downstream validations, such tools promise to accelerate our ability to positively impact human health and wellness.”

Related Links:
MIT