Skip Navigation


CIT can broadcast your seminar, conference or meeting live to a world-wide audience over the Internet as a real-time streaming video. The event can be recorded and made available for viewers to watch at their convenience as an on-demand video or a downloadable podcast. CIT can also broadcast NIH-only or HHS-only content.

Pi Day: Models and Data in Biomedicine: What's Real and What's Noise? And, Why Should We Care?

Loading video...

588 Views  
   
Air date: Monday, March 14, 2016, 1:00:00 PM
Time displayed is Eastern Time, Washington DC Local
Views: Total views: 588, (368 Live, 220 On-demand)
Category: Special
Runtime: 01:16:29
Description: Data Science Distinguished Seminar Series

If you think of a scatterplot of data overlaid with a model for the data and ask practitioners from different fields, “what’s noise and what’s real?” the answers may surprise you. To a biologist, the data will almost surely be “what’s real” and the model is a poor approximation to the “truth.” To a physicist, the model is probably “what’s real” and the data is just a noisy realization of an underlying true physical process that we are attempting to study. As we think about the biomedical data enterprise in the 21st century and the massive amounts of data we generate (and want to analyze!), we need to support multiple world views and have guidance on how to translate noisy data and noisy models into actionable information. Dr. Bustamante's presentation will draw upon several examples from Population Genetics (a field very rich in theory) and Genomics (a field not so rich in theory and much more data driven) to illustrate these points. It will also touch upon reproducible research and the question of how funding agencies need to support ecosystems for collaborative research including data producers, consortia, and so called "research parasites” that may want to use the data in ways that go beyond what the original experimental designers envisioned.

For more information go to https://datascience.nih.gov/PiDay2016/Schedule/Lecture
Debug: Show Debug
NLM Title: Models and data in biomedicine : what's real and what's noise? and why should we care? / Carlos Bustamante.
Author: Bustamante, Carlos.
NIH Pi Day (Conference)
National Institutes of Health (U.S.),
Publisher:
Abstract: (CIT): Data Science Distinguished Seminar Series If you think of a scatterplot of data overlaid with a model for the data and ask practitioners from different fields, "what"s noise and what"s real?" the answers may surprise you. To a biologist, the data will almost surely be "what"s real" and the model is a poor approximation to the "truth." To a physicist, the model is probably "what"s real" and the data is just a noisy realization of an underlying true physical process that we are attempting to study. As we think about the biomedical data enterprise in the 21st century and the massive amounts of data we generate (and want to analyze!), we need to support multiple world views and have guidance on how to translate noisy data and noisy models into actionable information. Dr. Bustamante's presentation will draw upon several examples from Population Genetics (a field very rich in theory) and Genomics (a field not so rich in theory and much more data driven) to illustrate these points. It will also touch upon reproducible research and the question of how funding agencies need to support ecosystems for collaborative research including data producers, consortia, and so called "research parasites" that may want to use the data in ways that go beyond what the original experimental designers envisioned.
Subjects: Data Accuracy
Genetics, Population
Genomic Structural Variation
Models, Biological
Publication Types: Congress
Webcasts
Download: To download this event, select one of the available bitrates:
[64k]  [150k]  [240k]  [440k]  [740k]  [1040k]  [1240k]  [1440k]  [1840k]    How to download a Videocast
Caption Text: Download Caption File
NLM Classification: QU 477
NLM ID: 101680583
CIT Live ID: 18658
Permanent link: https://videocast.nih.gov/launch.asp?19548