In the continuous fight against cancer, a new groundbreaking development has emerged from the collaborative efforts of researchers at Sanford-Burnham Prebys and the National Cancer Institute, part of the National Institutes of Health. They have introduced a pioneering artificial intelligence tool, PERCEPTION (PERsonalized Single-Cell Expression-Based Planning for Treatments in Oncology), which aims to transform the landscape of cancer treatment by predicting patient responses to various cancer therapies with remarkable precision.
Published in the esteemed journal Nature Cancer, the study highlights the uniqueness of each cancer case, noting that the disease’s complexity often hampers the effectiveness of one-size-fits-all treatments. Traditionally, cancer therapies have focused on targeting mutations in cancer-driver genes identified through genetic sequencing. However, these methods have often fallen short, with many patients not benefiting from such targeted approaches.
The new AI-based strategy, developed by Dr. Sanju Sinha, Ph.D., an assistant professor in the Cancer Molecular Therapeutics Program, and his team, including senior authors Eytan Ruppin, M.D., Ph.D., and Alejandro Schaffer, Ph.D., offers a fresh perspective by employing single-cell resolution in their analyses. This technique allows for a detailed examination of the transcription factors and messenger RNA molecules expressed by genes within each tumor cell, providing a comprehensive view of the tumor’s cellular composition and dynamics.
“Understanding the clonal architecture of a tumor at this granular level is crucial because it helps us track how cancer cells evolve and develop resistance to treatments,” explains Dr. Sinha. By employing a method known as transfer learning, the PERCEPTION model utilizes extensive pre-existing gene expression data and more limited single-cell data from cell lines and patient samples to refine its predictive capabilities.
The robustness of PERCEPTION was validated in three independent clinical trials targeting multiple myeloma, breast cancer, and lung cancer. In these trials, PERCEPTION distinguished between patients likely to respond to treatment and those not. Notably, in lung cancer, the tool was able to detect the emergence of drug resistance as the disease progressed—a breakthrough that could potentially lead to more adaptive and dynamic treatment plans.
Despite these promising results, PERCEPTION must still be prepared for clinical use. “The accuracy of our predictions depends greatly on the quality and quantity of the data we have,” Dr. Sinha notes, emphasizing the need for broader adoption and integration of such AI tools in clinical settings to gather more comprehensive data.
Dr. Sinha and his team are optimistic about the future of PERCEPTION and its role in clinical oncology. This tool could significantly improve treatment outcomes and offer new hope to patients facing this challenging disease by providing a data-driven, patient-specific approach to cancer treatment.
For more information on this transformative research, the entire study can be found in the latest issue of Nature Cancer under the title “PERCEPTION: Predicting patient treatment response and resistance via single-cell transcriptomics of their tumors.”
This development not only highlights the potential of AI in revolutionizing cancer treatment but also serves as a call to action for the broader medical and research community to embrace and advance these technologies, fostering a new era of precision oncology that is more adaptive, effective, and personalized than ever before.
Source: Medicalxpress
More information: PERCEPTION: Predicting patient treatment response and resistance via single-cell transcriptomics of their tumors, Nature Cancer (2024). DOI: 10.1038/s43018-024-00756-7
Like this article? Keep up to date with AI news, apps, tools and get tips and tricks on how to improve with AI. Sign up to our Free AI Newsletter
Also, come check out our free AI training portal and community of business owners, entrepreneurs, executives and creators. Level up your business with AI ! New courses added weekly.
You can also follow us on X