International Science Forum on Computational Toxicology

I. Virtual Tissues - the Next Big Step for Computational Biology. Implications for Toxicology and Risk Assessment: Relatively simple computational models of biological systems, best exemplified by PBPK models, have provided a new level of rigor for the analysis of the pharmacokinetic data and for our understanding of how pharmacokinetics influences dose-response. The "omics" revolution in the laboratory, and parallel developments in computer software and hardware, have moved the idea of virtual tissues (VT) from the realm of science fiction to forming the basis for a new, if ambitious, field of research. The potential payoffs from development of VT in toxicology are significant. VT development will build on current successes with PBPK modeling and take the development of quantitative descriptions of biological mechanisms to a new level of complexity. VT will have much greater capabilities than PBPK models for providing insights into dose-response and time-course behaviors and will promote inclusion of larger amounts of integrated biological data into risk assessment

II. Use of Computational Tools for Ecological Assessments: Molecules to Ecosystems: Ecological risk assessments need to focus on responses across multiple biological levels of organization. An understanding of processes at molecular, biochemical and cellular levels of organization enables insights into mechanisms underlying biological changes and, as such, provides a basis for extrapolation across species and chemicals. However, to make decisions about possible risk it is necessary to link alterations at these lower levels of organization to adverse effects in individuals and populations. An ultimate goal would be to understand how changes in populations of plants or animals affect specific ecosystems. This session will explore how computational approaches can be used as the basis for understanding and predicting effects of chemical stressors across the biological continuum of molecules to ecosystems.

III. Understanding Gene-Environment Interactions for Improved Risk Characterization: It is well known that different species and individuals within species react in different ways to identical exposures to environmental chemicals. This is in part driven by genetic variation in systems ranging from adsorption, transport, metabolism and receptor interactions. Understanding the mechanisms behind variable response can help determine vulnerable individuals and species. This can in turn drive testing protocols and (potentially) regulatory thinking. Additionally, performing toxicogenomic experiments in genetically heterogeneous populations can help to determine the connectivities in complex biochemical networks. Application of emerging tools in molecular biology is facilitating investigation of genetic contributions to consequences of environmental exposures. The objective of this session is to consider available tools and approaches for studying gene-environment interactions with a specific focus on improving environmental risk characterization.

IV A. Modeling Signaling as a Determinant of Systems Behavior: The goal of this session is to describe how biological signaling processes lead to overall systems behaviors (e.g., dose-response relationships). Modeling simpler and more complex systems provides insights into how different regulatory modules function. The modeling can provide insights into therelationships between the behaviors of parts of the system and the overall behavior when the parts are combined. Once the behavior of the biological system has been described, questions can be raised about perturbations of the system. Such perturbations can include altered physiological states (e.g., stress), disease states, and exposures to pharmaceutical or environmental compounds

IV B. Computational Modeling of HPG Axis: Over the past decade there has been a focused international effort to identify possible adverse effects of endocrine active compounds (EAC) on humans and wildlife. Effects on development, reproduction, aging, and hormone-sensitive cancers mediated through alterations in the hypothalamus-pituitary-gonadal (HPG) axis have been of particular concern. The development and application of computational models of the HPG axis can improve our understanding of the complex linkages between chemical exposures, biological dose, and effects of EAC to help predict the dose-response behavior and identify predictive biomarkers indicative of adverse effects. This session will focus on the development of computational models of the HPG axis at all levels of biological organization (i.e. intracellular, tissue, multi-organ systems), and the use of EAC to perturb the HPG axis to understand function and/or dynamic

IV C. Predicting the Environmental Fate and Transport of Chemical Contaminants: Exposure assessment requires knowledge of the environmental concentration and speciation of the chemical contaminant(s) of interest. This session will focus on: (1) the presentation of the computational tools currently available for predicting/simulating the fate and transport of chemicals in the environment; and (2) the identification of the most significant sources of uncertainty concerning our ability to predict chemical fate and transport.

IV D. Dose Response and Uncertainty in Risk Assessment: This session will examine some issues surrounding probabilistic dose-response assessments. While the Agency has some experience using probabilistic methods in risk assessment, virtually all of it is in using probabilistic methods to characterize uncertainty and variability in exposure. The goals of this session are to: (1) explore some methods for better characterizing the uncertainties inherent in estimating dose-response from toxicity data; (2) look at issues involved in extrapolating from animal to human dose-response: (3) look at approaches to dose-response assessment that probabilistically characterize uncertainty.

V A. Toxico-informatics: This session will cover the integration of chemical and toxicity data, toxicity data models, chemoinformatics, and the harnessing of legacy toxicity data to advancing predictive technologies

V B. Computational Molecular Modeling Applied to Understanding and Predicting Chemical Toxicity: This session will cover research modeling the modes and mechanisms for chemical toxicity from the viewpoint of the physico-chemical interactions between molecules in biological systems. This will deal primarily with a causal approach to this problem. It will also include advanced methods that increase the speed of these computationally rigorous applications so they may be used for screening.

V C. Application of Drug Discovery Technologies in Environmental Chemical Prioritization: Use of modern drug discovery technologies, including high-throughput biochemical and cellular assays, provides a new opportunity to survey environmental chemicals for potential for hazard. This session will focus on methods traditionally used in the drug discovery process for characterizing the bioactivity of small molecules with regard to target specificity and toxicity and how they can be applied to the field of environmental toxicology.

V D. Using Genomics to Predict Potential Toxicity: Genomics provides detailed molecular data about the underlying biochemical mechanisms of disease or toxicity, and could represent sensitive measures for detecting effects of environmental exposures. Thus genomics can provide useful data along the source-to-outcome continuum, when appropriate bioinformatic and computational methods are available for integrating molecular, chemical and toxicological information.