Using Artificial Intelligence to Detect Anti-cancer Immunity

UT Southwestern Medical Center and the University of Texas MD Anderson Cancer Center researchers developed an artificial intelligence method to determine anti-cancer immunity.

According to researchers, the pMTnet technique could create predictive analytic methods to conclude cancer prognosis and immunotherapy response.

“Determining which neoantigens bind to T cell receptors and which don’t seem like an impossible feat. But with machine learning, we’re making progress,” said senior author Tao Wang, PhD, said in a press release.

Mutation in the genome of cancer cells causes them to display different neoantigens on their surfaces. Some of these neoantigens are identified by immune T cells that search for signs of cancer and foreign invaders, allowing cancer cells to be destroyed by the immune system. However, some neoantigens seem invisible to T cells, causing them to grow unchecked.

“For the immune system, the presence of neoantigens is one of the biggest differences between normal and tumor cells,” said Tianshi Lu, first co-author and doctoral student in the Tao Wang lab.

“If we can figure out which neoantigens stimulate an immune response, then we may be able to use this knowledge in a variety of different ways to fight cancer,” Lu said.

With predictive analytics working to determine which neoantigens are recognized by T cells, researchers could develop personalized cancer vaccines, better t cell-based therapies, or foresee how patients respond to different immunotherapies.

Since there are tens of thousands of different neoantigens and methods to predict outcomes that could be time-consuming, technically challenging, and costly, researchers investigated the use of machine learning.

The team trained a deep learning-based algorithm named pMTnet using data from known binding or nonbinding combinations or three different components: neoantigens, major histocompatibility complexes (MHCs), and T cell receptors (TCRs).

The researchers tested the algorithm against a dataset developed from 30 different studies that had experimentally identified binding or nonbinding neoantigen T cell-receptor pairs, demonstrating that the new algorithm was accurate.

Researchers used the tool to gather information on neoantigens cataloged in The Cancer Genome Atlas and showed that neoantigens typically trigger a stronger immune response than tumor-associated antigens.

Additionally, it predicted which patients had better responses to immune checkpoint blockade therapies and better overall survival rates.

“As an immunologist, the most significant hurdle currently facing immunotherapy is the ability to determine which antigens are recognized by which T cells in order to leverage these pairings for therapeutic purposes,” said corresponding author Alexandre Reuben, PhD, Assistant Professor of Thoracic-Head & Neck Medical Oncology at MD Anderson.

“pMTnet outperforms its current alternatives and brings us significantly closer to this objective.”