Session 5: BEL networks in risk assessement

EnaBELing a Functional Ontology for Adverse Outcome Pathways – Facilitating Future Risk Assessment
Dr. Marja Talikka: EnaBELing a Functional Ontology for Adverse Outcome Pathways – Facilitating Future Risk Assessment
PRESENTING AUTHOR: 

Dr. Marja Talikka

INSTITUTION / COMPANY : 

Philip Morris International R&D

AUTHOR(S): 

M.Talikka, D Drubin, Jennifer Park, S Boue, J Szostak,
MC Peitsch, J Hoeng

>> SEE THE PRESENTATION <<

ABSTRACT CONTENT / DETAILS: 

Adverse outcome pathways (AOPs) provide a predictive, mechanistically-detailed framework to address the modern needs of toxicology, especially with the growing demands and inadequacies of current testing paradigms.

A unified biological ontology applied to this Systems Toxicology framework could facilitate its development and more widespread adoption, collaboration, and applicability across toxicological fields.

For this ontology, we propose the Biological Expression Language (BEL) as a tractable solution that is immediately adaptable to the AOP framework.

BEL is a triple-based language, where statements have a cause, relationship link, and an effect, very much like the conceptual structure of AOPs. BEL infrastructure can take the prior literature knowledge and apply it to data from a specific experimental system to identify impacted biology as well as to derive tailored AOP models.

In essence, BEL provides granular mechanistic detail, analytical tools, and extensibility by which to grow into the needs of the developing AOP concept. Here, we show a use case in which we have applied the BEL-based biological network models to analyze microarray data from the nasal epithelium of rats exposed to formaldehyde by inhalation.

Quantitative scoring of the short-term test data against BEL networks could serve as a predictor of long-term adverse outcome in the rat nasal epithelium.

In summary, biological network models built in BEL offer advantages from building to scoring AOPs for hazard and risk assessment.