Predictive Cancer Analytics

Cancer Research for Personalized Medicine

In collaboration with the Erasmus Medical Center and UMC Groningen, Finaps and SAS made predictions for cancer cases in order to move towards more personalized medicine. Solely on patient’s genetic information, we were able to make a prediction about whether a patient has a higher or lower chance of surviving cancer.

The Challenge

Our genetic data holds the key to understanding diseases, such as cancer, that originate from defects in the genome and it’s expression. Differences in genetics may also explain why some people respond better to certain treatments than others. Due to advances in sequencing technology it is becoming easier and cheaper to obtain genetic data from patients. However, leveraging this data to allow for the development of better treatments is a complex data analysis problem. The challenge was how to analyze genetic data from various modalities to gain insights on survival and response to therapy.

Cancer Research Solution

To approach this problem, the team from Finaps and SAS combined the various data sets and applied several steps of pre-processing. Next, we summarized the complex, high-dimensional genetic data into a much smaller representation. In this summary, we looked at whether patients with a similar summary had similar survival rates. This was found to be the case, which means that based solely on a patient’s genetic information, we are able to make a prediction about whether this patient has a higher or lower chance of surviving cancer. For a more detailed view of the approach, see the article on our News page. We anticipate that methods such as this one will find a way to the medical practice and will enable a more personalized approach in cancer treatments.


4 Data Scientists


SAS, Python


24 hours

Interested? Come have a cup of coffee with us!

Tim Pijl

Tim Pijl

Author, Data Engineer at Finaps