This week’s list of top data news highlights covers July 23, 2022 to July 29, 2022 and includes articles on installing artificial reefs to collect marine data and training an AI system to detect patient deterioration.
1. Predicting Protein Structures
DeepMind has expanded its freely available database of predicted protein structures to include almost every known protein. The database, which used to contain predictions for 350,000 proteins, now holds over 200 million protein structures. The company created and used an AI system to predict the structure of each protein.
2. Observing Reactions to Soccer
Manchester City soccer club has partnered with Cisco, a U.S.-based technology company, to create a smart scarf that uses sensors to observe fans’ physiological reactions to game activity. The scarf’s sensors can detect movement and changes in temperature, heart rate, and skin sweat.
3. Estimating Opioid Overdose Trends
Researchers at the U.S. Centers for Disease Control and Prevention (CDC) have created a machine learning model that can estimate how many people die from overdosing on opioids each week in the United States in close to real time. The team trained the system to use data from online searches, the National Forensic Laboratory Information System, emergency department visits, and Twitter and Reddit posts about heroin and synthetic opioids. With this data, public health officials can determine national trends more quickly, as data on opioid overdose deaths is often severely delayed.
The White House has launched a new website with GIS mapping data, real-time data on temperatures, and weather forecasts to help state and local officials better prepare for extreme heat events.
Researchers at Harvard and the University of Washington have created an AI system that can design proteins with specific functional properties. The team trained the system with data from the Protein Data Bank, a public database of protein structures. Medical researchers can use the system to create protein-based treatments or vaccines, such as COVID-19 vaccines.
6. Installing Artificial Reefs
South Carolina’s Department of Natural Resources has started installing modular reefs off the coast of the Charleston area. Once installed, officials plan to attach a buoy that can collect and transmit data on waves, water temperatures, and wind. The state plans to use the data to improve weather modeling and forecasting, and provide mariners with information about local conditions before sailing.
7. Identifying Mental Conditions
Researchers at Georgia State University have created an AI system that can detect signs of autism, schizophrenia, and Alzheimer’s disease from patterns in the brain before physical symptoms appear. The team trained the system with brain scans from 10,000 patients without any of the conditions and 1,200 patients with one of the conditions.
Researchers at the University of California, Santa Cruz have released new findings on the extent of the North Pacific Blob, a marine heatwave that lasted from 2013 to 2017. The team used data collected with sensors attached to elephant seals and found that subsurface warming existed at deeper levels than previously thought.
9. Detecting Patient Deterioration
Researchers at the University of Michigan have created an AI system that can detect abnormal or unstable blood pressure. The team trained the system with health data from over 5,000 patients. In tests, the system detected patient deterioration better than current monitoring methods.
10. Learning About Stormwater Systems
Officials in Culver City, California have partnered with Trigger XR, a U.S.-based mixed reality agency, to create an augmented reality project detailing the city’s stormwater management system. The project will include experiences such as visualizing underground stormwater systems, the history of native turtles, and images from the city’s history.
Image credit: Flickr user FWC Fish and Wildlife Research Institute