Unmet Need
There are more than 600,000 annual cancer deaths in the United States. One of the biggest hurdles in combatting and curing different types of cancer is understanding the etiology of cancer. Cancer is the end-result of a process of accumulation of driver epigenetic mutations and other events. It is well-known that some of these driver mutations may be inherited and that a relevant fraction of them may be due to negative effects of poor lifestyle choices or harmful environmental exposures. However, a third factor that is often overlooked is the random mutations accumulating during the normal process of cell division, independently of inherited and environmental influences. There is a need for a method or procedure to measure and track these random mutations to determine if they could be a significant contributor in the origin of certain cancers.
Technology Overview
Inventors reasoned that a supervised approach, an approach that takes in knowledge of the presence and intensity of an exposure in a cancer patient, would be a powerful way to identify mutational signatures of factors for which clinical and sequencing information is available. They analyzed the sequencing data of thirty-one types of tissues. The total number of each of the six possible point mutations types, as well as the totals for each of the ninety six possible combinations was collected for each of the patients along with their age at cancer detection. Only patients without any known environmental or inherited factors were considered. The data was then analyzed by the inventors new methodology, which is based on logistic regression in order to determine the mutational signature associated with age in each tissue. The use of logistic regression in the methodology enables the inventors to automatically test the method’s ability to predict and classify the age of each patient into two classes: young vs. old. This supervised method was about 82% successful, whereas the unsupervised methods have produced a 65% accuracy so far.
IP: Pending US application 17/616,740 (Published PCT WO2020247752A1).
Publication: Afsari B, Kuo A, Zhang Y, Li L, Lahouel K, Danilova L, Favorov A, Rosenquist TA, Grollman AP, Kinzler KW, Cope L, Vogelstein B, Tomasetti C. Supervised mutational signatures for obesity and other tissue-specific etiological factors in cancer. Elife. 2021 Jan 25;10:e61082. doi: 10.7554/eLife.61082.
