Written by Prashanti Manda
I attended the Pacific Symposium on Biocomputing (PSB) in Jan 2016. I presented a talk titled “Investigating the importance of anatomical homology for cross-species phenotype comparisons using semantic similarity.” This work explores the utility of including anatomical homology when computing semantic similarity of phenotype profiles. The majority of talks at PSB were focused on disease analytics and use of clinical phenotypes. There was a good balance of computer scientists and biologists at the meeting. An interesting session at the meeting was the social media session that was focused on large scale data analytics from sources such as Twitter and Instagram to track the spread of epidemics.
I also attended ICBO 2016 to present my work on the impact of annotation granularity on semantic similarity of phenotypes. I also served on the program committee for ICBO 2016 and was one of the poster judges at the conference. The title of my talk was “Measuring the importance of annotation granularity to the detection of semantic similarity between phenotype profiles”. Considerable human effort and time is invested to curate phenotypes in great detail from biological and medical literature using standardized ontologies. However, it is unclear if this level of detail is important for effectively measuring semantic similarity between phenotype profiles. In my work, I tested the statistical sensitivity of widely used semantic similarity metrics at varying levels of annotation granularity to test if higher annotation granularity improves the sensitivity of similarity metrics.
Attending ICBO gave me the opportunity to present my work to a diverse group of scientists focused on varying ontological applications. I found that the conference featured scientists and researchers from a wide range of areas within biology, medicine, ecology, computer science and text-mining. Of particular interest to me were the BioCreative sessions which focused on a variety of natural language processing and text mining applications to extract knowledge from scientific literature.
Lastly, I would like to acknowledge travel support from the Phenotype RCN for conference travel.