Research question: 

What tools can we use to better phenotype cancer symptoms and toxicities, and how do we implement them for research and clinical purposes?

​

Summary: 

As cancer survivorship increases, there is new emphasis on quality-of-life factors, particularly for symptom and toxicity management. The focus of this research is to develop a data science pipeline that integrates and analyzes cancer patient-reported and laboratory symptom/toxicity data. It combines the latest biostatistical and bioinformatic methodological approaches to generate appropriate phenotypes that accurately describe the complex nature of these variables.

​

​

Improved phenotyping of patient-reported symptom/toxicity measurement can improve methods for collecting such data routinely institution-wide, clinical patient management, and a wide range of research endeavours from genomic research to health services research. Symptom/toxicity data collection will only increase in clinical and research settings, with mounting evidence that symptoms/toxicities represent biologically-based outcomes. Therefore, proper phenotyping will be essential in clinical and research endeavours.

​

​

Scientific Goals:

(1) To create an automated symptom and toxicity data science integration pipeline that automatically pulls patient-reported symptom/toxicity accurately and completely

​

(2) To develop biostatistical and bioinformatic tools for identifying and predicting symptom/toxicity phenotypes

​

(3) To implement an analytical pipeline that demonstrates the utility of these phenotypes for clinical and research purposes