[Updated 24 July 2015: Datathon Finishes Strong with Team Presentations; added links to dataset webpages]
The Cross Neurodegenerative Diseases Datathon was held on June 30th through July 2nd
The Datathon focused on developing approaches to collaborate using data science to understand similarities and differences across different neurodegenerative diseases. About 30 scientists from several leading institutions from around the world participated in five teams. The teams had access to public datasets that are of interest to research teams studying Parkinson’s and other neuromuscular diseases. These were loaded into the tranSMART Platform. During the Datathon, the participating data scientists used the Platform and datasets to search for potential cross-disease biomarkers and to advance our understanding of disease pathophysiology that may inform new therapeutic approaches.
- What are the common biomarker changes across Parkinson’s Disease and Alzheimer’s Disease?
- What are the common pathway changes across Parkinson’s Disease and Alzheimer’s Disease?
- What is the normal distribution of imaging and fluid biomarkers across controls?
- Develop novel hypotheses or novel conclusions about biomarkers across the diseases and the studies.
Research Tools and Environment
The latest release of the tranSMART Platform v1.2 was available for the participants to actively use during the Datathon. In addition, several additional tools will be available including popular statistical, visualization, pathway and genomics viewers. Spotfire was made available by tranSMART Member PerkinElmer for use at the Datathon and they will have Spotfire experts on site. TranSMART Member IO Informatics was also present with their knowledge management tools. Thomson Reuters scientists also provided R expertise through the tranSMART R interface.
Public datasets of interest to the neurological disease communities were loaded along with several datasets. All of the large studies (ADNI, PPMI, LRRK2, and BioFIND) were curated by Thomson Reuters team and provided by the MJFF. The University of Luxembourg curated 10 GEO datasets and they provided their Parkinsons disease map to the Teams.
Summary of the Datathon Findings
Results from each team will be summarized here later this summer. Check back to see their results and conclusions.
Platinum Sponsor – PerkinElmer
The tranSMART open-source platform provides a proven infrastructure to assemble and centralize data and to analyze those data in pre-defined, web-based workflows. While those workflows support their intended analyses well, they lack the dynamic, interactive, and exploratory nature of a visual analytics tool like TIBCO Spotfire®. Moreover, any time scientists require a non-standard analysis, they have to enlist the aid of a data scientist or bioinformaticists that can custom build the requisite script. All too often those resources cannot be accessed before the idea is lost in the shuffle of day-to-day work.
With the Connector for tranSMART, scientists can pull tranSMART data into TIBCO Spotfire® as easily as they can find it and then employ the powerful and intuitive interface of TIBCO Spotfire® to develop better hypothesis more rapidly. Also, the Connector for tranSMART integrates with intuitive and powerful OmicsOffice® workflows, so the scientist can identify diagnostic biomarkers for complex biological processes at the click of a button. By connecting and combining these systems, the Connector for tranSMART enables the enterprise to leverage their investment in tranSMART and the scientists to spend more time analyzing the data.
Silver Sponsor and Host – Thomson Reuters
Dinner Sponsor – IDBS
IDBS is a leading global provider of advanced software for research and development (R&D) organizations to securely capture, manage, share and exploit data. Our technology and domain expertise enables users to link data to data, data to people and people to people to drive innovation, achieve faster time to market and improve margins. Our diverse customer list includes R&D driven international companies in pharmaceuticals, biotechnology, agricultural sciences, chemicals, consumer goods, energy, engineering, food and beverage, and healthcare. Founded in 1989 and privately held, IDBS is headquartered in the United Kingdom with offices across Europe, Asia and the United States.
Webinars to Prepare for the Datathon
- COMPLETED –
Monday June 22nd (10AM Eastern) – Special Training Class for Datathon Attendees (Thomson Reuters)
Tuesday June 23rd (10AM Eastern) – What is a Datathon; what will happen over the three days; what data will be available. (Keith Elliston, Ken Kubota)
- COMPLETED –
Wednesday June 24th (10AM Eastern) – Data resources from the U of Luxembourg including selected GEO Parkinson’s Disease (PD) datasets; Manually curated PD disease datasets loaded into tranSMART; a PD Map model in Galaxy; and the curation of the ADNI dataset.
Thursday June 25th (10AM Eastern) – Overview of Tools of various vendors including PerkinElmer (Spotfire) and IO Informatics.