Improving Situational Awareness in a High-Burden Labor and Delivery Suite in Malawi, Literature Based Similarity Score of Protein-Protein Interactions and LENS: Web-based Lens for Enrichment and Network Studies of Human Proteins
JoAnna Hillman, MPH, Doctoral Fellow
Title: Improving Situational Awareness in a High-Burden Labor and Delivery Suite in Malawi
Abstract: The most dangerous place to give birth is in Sub-Saharan Africa, where 62% of the world’s maternal deaths occur. The risk of maternal and neonatal death is highest during birth and the following day, making the intrapartum period critical for intervention. To reduce maternal deaths, the World Health Organization recommends facility-based deliveries by skilled birth attendants. The increasing demand for facility-based delivery strains the existing workforce beyond capacity due to severe health worker shortages, which jeopardizes the quality of obstetric services provided. In this proposed research, we will develop a model for a prototype informatics intervention that promotes regular patient monitoring, and provides situational awareness in a high-burden labor and delivery suite in Malawi, Africa.
Adam Roth, BS, Masters Fellow
Title: Literature Based Similarity Score of Protein-Protein Interactions
Abstract: Protein-protein interaction discovery and prediction are the key elements in aggregating and utilizing huge sums of important information which has been lost in the ever growing biomedical literature. Recent progress in biomedical event extraction has aimed to accurately capture the full semantics of a given sentence. Current systems build on binary relation extraction by capturing additional information about a given interaction such as direction and type. We have developed UPSITE, a text mining tool for validating interaction hypotheses. UPSITE is able to provide supporting information for interaction hypotheses by collating and distilling massive amounts of information which would otherwise remain frozen in free-text. Relation extraction performance was tested on the HPRD50 corpus and achieved an f-score of .88.
Adam Handen, BS, MS, Doctoral Fellow
Title: LENS: Web-based Lens for Enrichment and Network Studies of Human Proteins
Abstract: Network analysis is a common tool for the investigation of diseases and biological mechanisms. Current tools provide results in formats difficult to read by humans, or require additional downloads and plugins for visuals. The Lens for Enrichment and Network Studies of Proteins (LENS) is a web-based tool that requires no downloads or plugins that performs network analyses on provided gene sets. The tool creates a visual of the network, provides easy to read statistics on connectivity, and Venn diagrams with significance values for the network’s association with drugs, diseases, pathways, and GWASs.