SOCR News 2018 MNORC SOCR HAC Workshop
Contents
SOCR News & Events: MNORC-IBIC/SOCR/HAC Health Data Analytics Workshop
Logistics
- Date: Fri Oct 12, 2018
- Place/Time: 1-5 PM, 426 N. Ingalls (SNB 1250)
- Organizers: MNORC-IBIC, SOCR Team, HAC
- Registration: (space is limited to 25!) Please use this link to register for the training workshop. If there is sufficient interest, we may offer a live stream via BlueJeans.
- Format:
- Presentations: capabilities, resources, and expertise (6 x 15-min)
- Participant-led challenges, case-studies, template below, (20-30-min)
- Hands-on Consulting, Try-It-Now, apply to new data (120-min)
- Participants should bring laptops, and datasets, to try some of the resources hands-on at the training workshop
Presenters
- Ivo Dinov: SOCR Platform and Data Science and Predictive Analytics
- Alexandr Kalinin: SOCRAT, ML analytics
- Simeone Marino: Analytics, CBDA, DataSifter
- Nina Zhou: DataSifter, Biostats, Analytics
- Syed Husain: Data Dashboard, Viz/DVT, ML/BlueML
- Jerome Choi: Nutrition and Obesity Case-Study (mothers and newborns)
Background
- Michigan Nutrition Obesity Research Center (MNORC) and the Integrative Biostatistics and Informatics Core (IBIC)
- SOCR Website and Health Analytics Collaboratory website
- SOCR Navigators
- SOCR Datasets and Challenging Case-studies
- Electronic Textbooks: Probability and Statistics EBook, Scientific Methods for Health Sciences, Data Science and Predictive Analytics
- Hands-on interactive visualization of extremely high-dimensional data (learning module and webapp)
- SOCR News & Events
- SOCR Global Users
- Sprint 2018 SOCR Retreat, End-Of-The-Academic-Year, 12-2 PM on Wed, 4/18/18, in SNB 1250 with SOCR Spring 2018 Retreat Photos (UMich GDrive).
IBIC/SOCR/HAC Services
- Provide expertise in experimental design and modeling for preclinical, clinical and translational research studies that integrate clinical, molecular, neurobehavioral and other phenotype data.
- Provide guidance on the appropriate data architecture to enable integration and mining of data.
- Provide guidance and training in techniques and technologies to integrate and mine investigator generated or existing data sets.
- Assist investigators in the development of secure, Health Insurance Portability and Accountability Act (HIPAA)-compliant databases.
- Develop and promote the use of software tools for data visualization.
- Collaborate with other investigators, projects and centers to develop optimal data handling procedures and data housing systems, provide researcher friendly reports with suggestions for appropriate analytical tools.
Case-Study Template
Big Data is becoming ubiquitous. To examine complex health conditions, intricate biomedical phenotypes, and causal relations, advanced analytical techniques and powerful computational methods are necessary to ingest, harmonize, process, analyze and visualize large, heterogeneous, multisource, incomplete, multiscale, and incongruent datasets (DOI: 10.1186/s13742-016-0117-6). This template shows some of the characteristics that need to be provided prior to data interrogation. Each case-study should include the following components:
All Training Workshop Participants are encouraged to prepare and submit the the Workshop GDrive partition a Case-Study that represents a common data, visualization, analytical, methodological, processing, or interpretation challenge encountered in their clinical, basic or translational research.
- Title: Brief but descriptive case-study title
- Overview: A brief summary of the case-study
- Driving Challenges: List a set of 3-5 questions that have clear healthcare applications that might be addressed, or at least examined by, using the dataset
- Meta-data: Define all data elements, describe the dataset, data dictionary, data format, etc.
- Data: Package (e.g., as ZIP and share on GDrive, M+Box, etc.) the complete dataset. No PHI! The data could represent observational, derived, or simulated data. In general, to justify use of advanced analytics, the case-study should represent a real and interesting phenomena (e.g., include at least 10 variables, one or more time-points and represent 100 + cases/subjects/instances, hopefully, hundreds or thousands of cases)
- Provenance: Include appropriate, references, URLs, PMCIDs, comments, credits, etc. describing the provenance of these data
Examples of many case-studies are available on the SMHS Case-Studies Canvas Site.
References
- This workshop is sponsored in part by NIH Grant P30 DK089503 and NSF Grants 1734853 and 1636840.
- SOCR Home page: http://www.socr.umich.edu
- GDrive
Translate this page:
