A new way to predict breast cancer survival
Researchers at Columbia University win a breast cancer prognosis challenge with a new computational model, which could potentially improve models for diagnosing and treating many types of cancer.
A researcher who used to work in digital television has just led a team of Columbia University engineers to win the Sage Bionetworks / DREAM Breast Cancer Prognosis Challenge.
Dimitris Anastassiou, who is now a systems biologist (meaning he investigates interactions within biological systems), reports in the April 17 issue of Science Translational Medicine that his team's winning computation model is extremely predictive of breast cancer survival.
Before the challenge, Anastassiou and his team identified what they call "attractor metagenes," which are genetic signatures expressed in almost the exact same way across many types of cancer. Their new computation model is based on examining three of these metagenes combined in very specific ways. This approach scored highest in the crowd-sourced challenge that used molecular and clinical data (including some 2,000 breast cancer samples).
"These signatures manifest themselves in specific genes that are turned on together in the tissues of some patients in many different cancer types," Anastassiou said in a school news release. "If these general cancer signatures are useful in breast cancer, as we proved in this challenge, then why not in other types of cancer as well? I think that the most significant -- and exciting -- implication of our work is the hope that these signatures can be used for improved diagnostic, prognostic, and eventually, therapeutic products, applicable to multiple cancers."
While it may seem unrelated, it is no coincidence that Anastassiou also holds patents used in the international standards of all types of DVDs; his work as a systems biologist may similarly lead to a meta model that can be applied to all -- or at least many -- types of cancer.
"The hallmarks of cancer are unifying biological capabilities present in all cancers, as described in some seminal papers," Anastassiou says. "We think that we have now reached the point where systems biology can also identify such hallmarks."
Today, doctors lean heavily on biomarker products that look at specific genes in cancer patient biopsies to determine which treatments are most appropriate for their patients. Anastassiou says it's worth investigating whether his precise "pan-cancer" signatures could actually improve the accuracy of these biomarker products.
The Sage challenge included 354 participants from 35 countries who openly submitted their models and were encouraged to incorporate other models into their own. In total, the contestants submitted more than 1,700 models, with Anastassiou's pan-cancer approach proving to be the best predictor of breast cancer survival.
Oh, and those patents from Anastassiou's previous work in digital television? They helped fund the breast cancer model research.