Epigenetics and Machine Learning Help Make Cancer Predictions
Did you know that NCI-funded researchers are using genetic data from tumor cells to forecast the future? Although many of today’s models are relying on genetic data to make their predictions, this model from researchers at the University of California-Los Angeles (UCLA), is different. This model looks at factors that blend both environmental and genetic influences (i.e., your epigenetics), to predict cancer outcomes.
Epigenetic changes can be a product of your environment, (e.g., smoking, sun exposure, or other lifestyle choices), or they can stem from internal factors (e.g., infection and inflammation). Your epigenetic patterns are unique to you. This makes them particularly promising for precision medicine, in which clinicians tailor treatment that best fits each patient.
The model offers some important insight into the proteins that make epigenetic changes (i.e., the “epifactors” that occur in a broad range of cancer types). According to first author, Michael Cheng, their model shows that epifactors have a role in how tumors progress, and these epifactors are useful in predicting outcomes across many different types of cancer.
Those findings have clinical implications. A model like this can help both in predicting patient survival and in targeting genes to disrupt cancer’s progression.
The study’s co-senior author, Dr. Hilary Coller, said, “Our extensive and unbiased survey of 720 epifactor genes helped us discover several novel genes that could be possible drug targets. We also identified signature epifactors, such as those associated with enzymes, as well as two protein families, that could be potential targets for epigenetics-based cancer therapy.”
You’ll be able to learn even more about epigenetic influences in the future. As per Dr. Mithun Mitra, co-senior author, “the research team will be extending their analyses to better understand these prognostic epifactors and how they impact cancer’s progression and treatment.”