The TPC pipeline measures 43,000 data points per tumor sample. To date, less than 200 data points are used for the TPC tumor report, and the majority of the collected data remain unexplored.The TPC Data Science track is developing new computational models for in-depth analysis of patient data to identify novel disease characteristics.
The ultimate goal is to untap this data source to guide personalised cancer treatment.
Key milestones that are pursued are the discovery of
- novel biomarkers
- mechanisms of resistance to standard treatment and
- novel therapeutic targets
To achieve this, data science and machine-learning algorithms will be used to investigate the expression level of markers as a function of various clinical variables, including new molecular markers and cell populations as a predictor of response to treatment.
Both the molecular data provided by TPC technologies and the patients’ clinical follow-up data (such as data on cancer treatment, side effects and response) will be fed into the computer models as inputs.
The comparative analyses of these multi-level datasets may lead to the identification of potential mechanisms of resistance to standard treatment and thus enable the discovery of new therapeutic targets.
Each of these findings will drive the formulation of novel hypotheses that can be tested in future samples and in clinical trials, as well as in publicly available datasets in biobanks to identify potential prognostic and predictive cancer biomarkers.