Breakthrough In Cancer Treatment: The Role of Generative AI In Drug Development

    Published on:

    We’ve all seen how generative AI tools like ChatGPT and Stable Diffusion can create amazing text and images that closely resemble those created by humans.

    But did you know it has many other potential use cases, including potentially helping treat cancer?

    Thanks to the speed and power of machine learning, it can be used to analyze huge datasets of compounds, drug trials, and clinical results very quickly. This allows researchers to speed up the difficult task of narrowing down test candidates to a final candidate.

    Demonstrating this is an Oxford-based biotech company Such recently achieved a world first in creating an immunotherapy drug with the help of generative AI.

    Let’s take a look at what this breakthrough means and how generative AI can become a key weapon in the fight against cancer.

    Immunotherapy and cancer

    Some cancer treatments, such as chemotherapy, work by attacking the cancer directly, while immunotherapy tries to strengthen the body’s defense system to more effectively destroy cancer cells.

    Etcembly’s research focuses on a type of immunotherapy drug known as a T-cell engager. T-cell engagers are designed to bring immune cells present in white blood cells closer to cancer cells and do their job of killing them.

    The project, known as ETC-101, accomplishes this by targeting PRAME (Preferred Expression Antigen in Melanoma), a protein frequently found in cancer cells but rare in healthy tissue.

    The idea is that the body’s natural defenses can be directed to the areas of greatest impact while minimizing damage to other parts of the body. This means fewer side effects and hopefully faster recovery time for patients.

    How was generative AI used?

    Etcembly has created its own generative AI engine called EMLy, based on Generative Large-Scale Language Models (LLM), the same technology that powers tools like ChatGPT.

    EMLy was used to scan the genetic code of T cell receptors. T-cell receptors are molecular mechanisms in the body’s immune system cells that help detect foreign or abnormal entities, such as viruses or cancer cells. By scanning hundreds of millions of these codes, we can determine which cells are most effective against specific cancer cells. In particular, we look for candidate leukocytes that are most likely to have a low pM affinity, i.e. cells that are most likely to form bonds with and destroy cancer cells.

    It also looks for cells that are less likely to harm nearby healthy cells, meaning they are less likely to cause harmful side effects.

    One way to imagine this is that every cancer cell has a unique key, and white blood cells are like a stack of keys. If you find a key that fits the lock, you can destroy the cell. But unfortunately, there are hundreds of millions of keys to try.

    Even traditional computer simulations take a long time to sort through all possible combinations, sometimes longer than the patient’s time. However, with generative AI like EMLy, new TCR sequences can be quickly created and tested in simulation based on all existing TCR and patient outcome information held in the training data.

    What’s next for generative AI in immunotherapy drug discovery?

    Research that combines generative AI with the search for immunotherapeutic treatments could accelerate the development of personalized medicine. These are treatments that are specifically tailored to each individual patient and are created by determining the TCR sequences that are most likely to be effective against an individual’s specific cancer cells.

    Backed and supported by Nvidia’s Inception startup incubator, Etcembly is currently seeking partners to collaborate with to develop more innovative treatments using EMLy technology.

    There are certainly ethical considerations to consider. Biases in training data can lead to less effective treatments for certain groups of people, which can lead to disparities in healthcare outcomes. Additionally, these methods rely heavily on personal data, including personal genetic data, and therefore require close attention to data privacy and security.

    However, demonstrating the potential of integrating LLM technology with complex medical drug discovery will accelerate the transition of breakthrough therapeutic candidates from the laboratory to the clinic, ultimately improving cancer survival rates. It is expected that. It also has the potential to dramatically reduce costs, which are often an impediment to drug discovery.

    In the near future, we may see breakthroughs in using generative AI to create compounds to treat many other conditions. And this trend will only accelerate as generation algorithms become more powerful and sophisticated over time.


    Leave a Reply

    Please enter your comment!
    Please enter your name here