Feb 3, 2018
New tools for pooling and analyzing multisite datasets
How can one efficiently combine experimental and observational data from different laboratories into a single predictive model when the laboratories have differently calibrated measuring instruments and the study populations have different demographic distributions? When both differences exist in the data, important quantities may not be identifiable. In a recent PNAS article, CPCP researchers provide sufficient conditions for determining when such quantities are identifiable and practical plans to account for the differences. The methods have been applied to two Alzheimer’s disease studies, where measured cerebral spinal fluid and demographic data show differences. These tools to add to the armamentarium of the scientific experimenter and data analyst for efficient combination of information from diverse sources.
Aug 15, 2017
High-throughput computing and machine learning boost drug screening
High throughput screening is a powerful, albeit expensive, method of drug discovery. The process can be made more cost effective by first using docking programs to do virtual screening of candidate drugs. In recent research, a CPCP team has shown that virtual screening can be made significantly more effective by coupling high-throughput computing, machine-learning, and mixture modeling methods to integrate the results of multiple docking programs. Read a UWMadScience story about the work here: https://uwmadscience.news.wisc...
Jun 2, 2017
CPCP sponsors Big Data Analytics-as-a-Service workshop
The first workshop on Big data analytics-as-a-Service: Architecture, Algorithms, and Applications in Health Informatics is taking place on August 14, 2017 (in conjunction with KDD 2017) in Halifax, Nova Scotia, Canada. The workshop will consist of a combination of invited keynote speakers, panel discussion, and paper/poster presentations. http://bigdas.org/
Oct 27, 2016
CPCP at the Wisconsin Science Festival
CPCP faculty, postdocs, staff, and graduate students worked together to develop and present three interactive Big Data exploration stations at the 2016 Wisconsin Science Festival. The Wisconsin Science Festival is a statewide event that strives to inspire curiosity and creativity by providing opportunities for people of all ages to explore, engage and discover science. This year the Wisconsin Science Festival event at the University of Wisconsin-Madison was attended by a large number of community members in the Madison area including more than 4,000 children on school field trips.
This outreach provided an opportunity for visitors to discover several types of Big Data studied in CPCP and to interact with some of the scientists working in the center. Visitors also had the opportunity to learn about the challenges faced by CPCP researchers as they work to develop new methods of analyzing biomedical Big Data with the goal of improving human health. At our neuro-imaging exploration station, visitors explored MRI images of the brain that CPCP researchers are using to predict Alzheimer’s disease. Our EHR exploration station used a set of Pokémon-based games to demonstrate to visitors the challenges facing researchers as they study electronic health records. At our final exploration station, visitors participated in a crowdsourced assembly of a human gene sequence using colored beads to represent the bases in DNA sequences. This activity provided an opportunity to discuss the Big Data challenges faced by researchers in CPCP’s Epigenome- and Transcriptome-based Phenotyping projects.
Aug 18, 2016
UW2020 WARF Discovery Initiative Grant Awarded
The CPCP Phenotype Models for Breast Cancer Screening project team was recently awarded a 2-year, $500,000 grant to advance the precise targeting of breast cancer prevention and early detection by rapidly moving from the discovery of individual DNA variants called single nucleotide polymorphisms (SNPs) that appear to predict the risk of developing breast cancer, to their application in clinical settings. The group will also add to the predictive power of SNPs by pairing them with information on an observed physical difference --breast density -- that captures the effects of both genes and the environment.
This UW2020 WARF Discovery Initiative grant is funded jointly by the Office of the Vice Chancellor for Research and Graduate Education and the UW Carbone Cancer Center.
Jul 26, 2016