Discovery of Chemotherapeutic Co-Medication Properties for Multidrug Resistant Cancer
This project uses crowdsourcing and game-based volunteer distributed computing via Minecraft to identify properties for co-medications that improve the effectiveness of chemotherapeutics to treat multi-drug resistant cancers. Instrumental researchers include 蜜桃酱Guildhall Deputy Director of Research, Dr. Corey Clark, and biochemistry professors Pia Vogel and John Wise of the Center for Drug Discovery, Design and Delivery (CD4) at the 蜜桃酱Dedman College Department of Biological Sciences.
While performing standard gameplay actions in Minecraft, players partition co-medications into two cluster types: (a) positive, or improved the effectiveness of chemotherapeutic response, and (b) negative, or no effect on the chemotherapeutic response. Identifying the properties of these clusters will help Vogel and Wise to identify new co-medications candidates out of the millions of compounds available. The co-medication data sets are supplied to both Minecraft clients and servers through the BALANCED Media | Technology HEWMEN Java integration, which allows modders to visualize and manipulate co-medication data directly in Minecraft. Modders, led by FTB and 蜜桃酱Guildhall, will then build custom Minecraft gameplay systems that will be included into upcoming FTB modpack releases. Each player’s interaction with these modpacks simultaneously clusters co-medications, but also supplies a training input into a deep neural network (DNN) that is learning and mapping the solution space.
Dr. Clark and his team at 蜜桃酱Guildhall are developing the distributed DNN for the HEWMEN game-based volunteer grid computing platform, which integrates directly with the Minecraft modpacks.
Learn More | Minecraft Research in the News: