Practical use with major implications: Applied Semantic Knowledgebase (ASK™) for Predictive Biology

Dr. Erich Gombocz

IO Informatics, Inc.


Abstract: While predictive biology has been a major goal in the development of safer drugs for quite a while, its promises have — despite major advances in analytics — been challenged by 2 major factors: the difficulties of meaningful semantic integration of heterogeneous experimental and public data, and the difficulty in understanding the complex biological functions involved.

Building on advanced data access and integration capabilities, Sentient applies semantic patterns to create conclusive network models using virtually any combination of internal experimental data and / or external published information. These patterns apply semantic SPARQL query technology to build complex searches across multiple information sets. SPARQL is a semantic search technology capable of detecting patterns within and between different data types and relationships, even if the initial datasets are not formally joined under any common database schema or data federation method. Such patterns or models are then placed in an Applied Semantic Knowledgebase (ASK) which is unique to a specific research focus, providing a collection of specific models applicable to screening and decision making.

This talk describes what is involved in using this technique for toxicity assessment based on multi-modal datasets. Typically, while many of the observed changes in metabolites, gene expression and tissue analytics can result from the same treatment, most likely they represent very different biological processes. In the liver, for example, metabolites might reflect changes in the urea cycle and intermediary metabolism, while gene expression may quantify alterations in the immune response and signaling pathways. As such, despite being pharmacodynamically correlated, the effects may not be directly functionally linked within the biological system. Thus, for biomarker discovery in predictive toxicology to be successful, it is mandatory to move from the bare collection of observed statistical correlations towards an integrated relationship network which helps differentiating the effects and by doing so understanding the biological functions involved. Generating test models which can be qualified, refined and validated for broader use has been previously challenging and required considerable efforts.

In a practical demonstration using combinatorial biomarkers to categorize treatment-specific responses involving a set of very different biological functions, the use of ASK is demonstrated. Insights gained from semantically joining coherent findings despite their different methodologies allow researchers to better understand how the organism responds to experiment-induced system changes and to recognize mechanistic aspects of biomarkers at a functional level.

Using ASK makes it possible to actively screen previously disconnected, distributed datasets, to identify and stratify results, applicable to target assessment, compound efficacy, toxicity profiling, disease signatures, clinical trials pre-screening, and patient stratification to provide informed decisions in life science and personalized medicine.

About the Speaker: Dr. Erich Gombocz has over 25 years of experience in Life Science research, laboratory automation and data management in scientific and distributed systems environments, plus more than 20 years programming experience in instrumentation control, user interface, database design, scientific analysis, and on-line laboratory automation as well as being developer of innovative software algorithms and architecture.

Focusing on semantic data integration and knowledge management in life sciences, he founded IO Informatics in 2003 together with Bob Stanley to apply systems biology approaches to challenges in the area of pharmaceutical and clinical decision-making.

Dr. Gombocz has published over 60 scientific publications and holds currently more than 40 biotechnology- and software-related US and international patents. He is an international expert in separation science and bioinformatics, a member of several professional organizations, and serves on the editorial board of a number of scientific journals.