Breakthrough in personalised medicine thanks to AI

A team of researchers at the University of Waterloo has developed an artificial intelligence program called GraphNovo that could lead to more personalized and effective cancer treatments. This new technology analyzes the molecular makeup of cells with much greater accuracy, allowing doctors to better understand how cancer differs between patients.

GraphNovo focuses on analyzing data from peptides, which are short chains of amino acids that act as building blocks within our cells. Peptides play a huge role in regulating bodily functions and identifying foreign or irregular cells. When cell mutations occur, the peptides in cancerous cells undergo changes that distinguish them from healthy cells.

The key to more personalized cancer treatments lies in mapping out these peptide differences unique to each patient’s cancer. As PhD candidate Zeping Mao explained, “What scientists want to do is sequence those peptides between the normal tissue and the cancerous tissue to recognize the differences.” However, quickly and reliably sequencing peptides in unfamiliar cancer cells has proven extremely difficult.

GraphNovo steps in

The AI program takes data from mass spectrometry a technique which can be employed to rapidly break down peptide samples from cancer cells and analyzes their amino acid sequences. It then employs advanced machine learning algorithms to fill in any missing data in this mass spectrometry data and accurately reconstruct the full peptide chains.

GraphNovo achieves far greater precision in identifying peptide sequences than past sequencing methods. By reconstructing the peptides found in a patient’s cancer cells, doctors gain crucial insights into how that specific cancer differs from previous cases at a molecular level. These insights can then guide more tailored treatment plans.

The possibilities for impact are staggering. As Mao stated, “If we don’t have an algorithm that’s good enough, we cannot build the [personalized] treatments.” More accurate peptide sequencing paves the way for immunotherapies that retrain each patient’s immune system to target their unique cancer cell markers. It also promises to enhance everything from early cancer detection to the development of new chemotherapy drugs.

Beyond cancer, accurately cataloguing peptides has major implications for the study of other diseases too. The approach used by GraphNovo could accelerate vaccine design for viruses like COVID-19 and Ebola by revealing invaluable molecular clues about how these pathogens function and spread.

While the practical medical applications are still in development, the University of Waterloo’s groundbreaking research underscores how artificial intelligence and machine learning are transforming biomedicine. Programs like GraphNovo are enabling scientists to unlock troves of data within cells that have never been accessible before.

“Soon, we will be able to use it in the real world,” said an optimistic Mao, highlighting the immense life-saving potential. Just as the first microscopes fundamentally changed our understanding of cells centuries ago, this new computational “microscope” promises to launch a new era of precision medicine in which treatment is tailored to each patient at the most fundamental level.

You can read the full paper here in Nature Machine Intelligence.

Staff Writer

Our in-house science writing team has prepared this content specifically for Lab Horizons

Leave a Reply

Your email address will not be published. Required fields are marked *