Microsoft, the tech giant synonymous with computers and computing, recently announced that they were getting involved in the cancer business. More specifically, Microsoft stated that they were going to use computer science to crack the cancer code within the next decade – a somewhat lofty goal.
According to Microsoft they are employing two over-arcing approaches in their quest. One of these methods is based on the idea that cancer and other biological processes are basically information processing systems. As such, researchers at Microsoft believe that they can successfully employ computational methods used to model and reason computational processes to model and reason about biological processes. The second approach is to employ sophisticated analysis techniques, such as machine learning, to the vast amounts of biological data that is currently available to better understand and treat cancer.
To support their research, Microsoft opened a wet laboratory over the summer of 2016, which they have staffed with a team of biologists, programmers, and engineers from around the world. These researchers have embarked on a number of interesting concepts and ideas in their effort to solve the problem of cancer.
One particularly intriguing approach is their so called “moonshot” idea; the design of tiny molecular computers synthesized from DNA, which will live within healthy cells and monitor them. When these molecular computers detect cells changing and becoming cancerous they can reprogram the cells, essentially rebooting the system to remove cancer cells.
Another approach that Microsoft is employing is called “debugging the system”. This method involves understanding the fundamental biological processes that lead to the onset of cancer in order to fix the problem. The researchers at Microsoft are attempting to understand the complex instructions or commands a cell uses to behave in a particular way. The goal is to compare the behaviour of healthy cells with unhealthy cells in order to develop a way to repair them. To do this the researchers use a cloud-based system to model how cells communicate and interact with each other called the Bio Model Analyzer, or BMA. Currently, Microsoft and AstraZeneca have been using the BMA to identify why leukemia patients respond so differently to treatments. Microsoft believes that one day the BMA will play a significant role in personalized medicine allowing oncologists to treat cancers on a person by person basis. For example, an oncologist may be able to look at other non-cancer medications that a patient might be taking to determine what sort of drug interactions may occur and modify treatments as required.
However, according to Microsoft, while they believe the BMA will revolutionize cancer treatment, it will only work if biologists are comfortable using it. Thus, the developers at Microsoft have spent years designing a system that is familiar and recognizable to biologists, and uses language that biologists would understand.
One other project the researchers have been working on is the development of artificial intelligence to evaluate 3D scans of a patients tumor pixel by pixel to ascertain how much a tumor may have changed since last scanned. Currently CT scans can produce up to 2000 images, all of which contain information, some of which is imperceptible to the naked eye. Using artificial intelligence could streamline the process and help radiologists make more informed decisions.
What Microsoft is proposing to do is an ambitious undertaking, however, if successful it could greatly improve our understanding of cancer and allow informed treatment decisions to be made on a patient by patient basis.