Auransa, Polaris harness AI, quantum computing to discover cancer drugs
Auransa and Polaris Quantum Biotech first teamed up last year, bringing together Auransa’s artificial intelligence (AI) capabilities and Polaris’ quantum computing (QC) based drug design platform to tackle neglected women’s diseases. Triple-negative breast cancer remains an unmet medical need despite efforts to tackle the aggressive tumor type.
Auransa and Polaris applied their capabilities to the challenge of finding a more effective treatment for the condition. Using Auransa’s SMarTR Engine and human disease data, the partners have identified a biological pathway and two protein targets that may be implicated in the cancer.
“From the molecular data from breast cancer tumors alone, without referring to any literature, our SMarTR Engine has correctly identified clinically proven hormone targeted therapies such as letrozole, tamoxifen, and others to be effective for the breast cancer subtypes that are estrogen receptor-positive. Similarly, the SMarTR Engine has correctly identified the target for herceptin as one of the top targets for Her2+ type breast cancer. Borrowing from the strength of these validations, we have confidence in the pathway and the two targets that the SMarTR Engine predicted for triple-negative breast cancer,” said a spokesperson for the companies.
Having identified the two targets, the collaborators used Polaris’ Tachyon platform to design and search a virtual chemical space containing billions of molecules. The work led to the identification of several sets of leads for each target that will undergo laboratory testing.
The technologies enabled the project to go from data collection to the identification of molecular leads in six months, according to the companies. Tachyon supported the timeline because quantum computing systems can perform optimization problems at greater rates than classical high-powered computing.
“Our Tachyon platform utilizes this strength to be able to search a chemical space that numbers in the billions for molecules that fit a very specific set of properties that identify the pharmaceutical target, and then quickly determines the most likely candidates. The ability of a quantum computer to quickly examine all of the potential solutions to a given multi-variable optimization problem in minutes rather than weeks is a very significant early advantage for the technology, and with projects like ours the application of this kind of powerful computing holds the potential to be world-changing in many industries,” said the spokesperson.