CROWD-POWERED HYPOTHESIS GENERATION AND TESTING IN MEDICAL AND HEALTH SCIENCE
Clinical research is difficult because no two patients are exactly alike. Even patients with the same disease have different external and internal factors that drive the progression of diseases. For instance variety of genomic predisposition, different immunities based on unique exposure histories, and different microbial biomes. Patients may also have varying clinical presentations, different co-morbidities, and comorbities. When analyzing clinical data, these phenotypic variables can be confounding, leading to false conclusions and findings that do not generalize to all patients. Further complicating matters are variations in lifestyle, mental health, diet, and unrelated drug therapies.
Thus, a patient's clinical picture comprises a unique, large, multidimensional data set, which often includes more variables than the number of patients in a typical study population. While traditional statistical methods can sometimes ferret out the influence of one or more independent variables, deciding which combination of variables should be studied in the first place is left to the purview of scientists. Unfortunately, this is very limiting to scientific discovery because the possible relationships among clinical variables that could be studied is extremely large relative to the number of scientists studying them.
But what if the general public, including patients themselves, had a way to identify and test possible relationships among clinical variables? What if they had a way to pursue promising avenues of inquiry, which did not require scientific or medical training? Furthermore, what if they could collaborate with other participants around specific biomedical topics, building on each others' results?
We propose to build an online platform for crowd-powered scientific discovery using clinical patient data. This platform will provide an intuitive visual interface for exploring clinical data, tools for selecting clinical variables to analyze, and collaboration mechanisms for sharing and discussing results. Such a platform would allow, for example, a patient with Chronic Fatigue Syndrome (CFS) to go online, and in five minutes test the hypothesis that aspirin is most effective for relieving fatigue when chronic pain is a comorbidity and then to share any noteworthy findings with the public, the CFS patient community, and with professional scientists for possible further investigation.
QUICK FACTS ON MESEARCH
- Scientific discovery in health and medical science is limited to disciplinary silos
- In the age of Big Data, open clinical data is largely available and can be used for participatory discovery strategies
- Large-scale participatory systems can leverage on human capabilites to decrease scientists lab work and accelerate scientific discovery
- Patients level of suffering fuels motivation to find solutions for their medical condition
- We will build a platform that helps patients, researchers and society.