Section 4: EPA's CompTox programs

Dr. Robert Kavlock, EPA Deputy Assistant Administrator for Science, talks about EPA’s new computational toxicology research programs. Click here to watch the full video introduction. Photo by EPA

In 2005, the Environmental Protection Agency (EPA) established the National Center for Computational Toxicology (NCCT) within the agency’s main research arm, the Office of Research and Development (ORD organizational chart). NCCT coordinates the agency’s computational toxicology research program, nicknamed CompTox. This page will briefly review the various research efforts within the CompTox program.

CompTox encompasses several discrete research programs, each focused on the development and application of innovative approaches and tools to improve chemical hazard, exposure and risk assessment. ToxCast is one such program. You can read more about the ToxCast program in Section 3 of this primer.

EPA created CompTox in large part to address the significant lack of health and environmental data on the thousands of chemicals for which it is responsible to ensure are safe for human health and the environment. EPA states that the purpose of the CompTox program is to “conduct innovative research that integrates advances in molecular biology, chemistry, and computer science to more effectively and efficiently rank chemicals based on risk.”

The individual research programs within CompTox are generating massive amounts of data on thousands of chemicals. Sophisticated computer systems and programs are required to integrate and analyze these large data sets. This is why the term “computational toxicology” is used to describe these newer types of chemical assessment tools and approaches.

For more information on CompTox visit:

The following subsections describe the individual research programs within CompTox.

Toxicity Forecaster (ToxCast)

The ToxCast program is focused on the development and use of high-throughput (HT) in vitro testing to identify and characterize chemical toxicity. HT testing and the ToxCast program are described in detail elsewhere in Section 2 and Section 3 of this primer.

For additional information from EPA on the ToxCast program visit:

Tox21 Program

The Tox21 program is an inter-agency program of which EPA is apart and to which ToxCast data are contributed. Tox21, established in 2008, is a collaborative effort that leverages federal resources and expertise from EPA, National Institutes of Environmental Health Sciences (NIEHS), National Institutes of Health (NIH), and the Food and Drug Administration (FDA) to conduct high-throughput in vitro testing for screening and prioritizing thousands of chemicals for potential toxicity.

The technology and intent of the Tox21 program are very similar to that of ToxCast. The primary differences are:

  1. There are far fewer assays being used currently in the Tox21 program. Tox21 contains approximately 50 of the 600 assays used in ToxCast.
  2. Tox21 is testing many more chemicals than ToxCast. The Tox21 program includes 10,000 chemicals, whereas ToxCast includes about 2,000 chemicals.
  3. Tox21 assays are conducted in house at the NIH National Chemical Genomics Center, whereas ToxCast assays are conducted by contracted biotech companies outside of federal agencies or research labs.

For more information on the Tox21 program, please visit:

Exposure Forecaster (ExpoCast)

The ExpoCast program is the exposure counterpart to ToxCast. ExpoCast is focused on developing and validating computer models to estimate human and environmental exposures to thousands of chemicals. To build the models, ExpoCast is using exposure information from the Centers for Disease Control’s NHANES biomonitoring program and EPA’s Toxics Release Inventory and Chemical Data Reporting system (formerly Inventory Update Rule), as well as available environmental and biological fate and transport data on chemicals. Notably, however, information on chemical exposures from consumer products is currently lacking in ExpoCast.

For more information on ExpoCast visit:

Virtual Tissues (v-Tissues)

A physiologic lobule

The virtual liver (v-Liver) project, part of EPA’s CompTox virtual tissues program, is one of several computer-generated models being developed to predict and evaluate chemical toxicity. Photo by EPA

The virtual tissues program seeks to develop computer simulations of complex biological structures and cell networks to predict how chemicals can affect and perturb these systems. The term in silico is used to describe these types of computer-based testing systems. EPA is focused currently on developing computer models of the liver and developing embryo, called v-liver and v-embryo, respectively. The v-embryo program is further divided into four focus areas: eye development, limb development, vascular system development, and embryonic stem cell development.

Developing computer simulations of these complex biological systems involves using advanced computer programming to overlay and integrate a wealth of scientific knowledge on biological pathways and processes underlying liver function and embryonic development. The predictive capacity of in silico testing systems like v-liver and v-embryo hinges on the extent and quality of data on which they are built. EPA is using both existing information from the scientific literature and new data generated from research programs like ToxCast to build these models.

In silico models are the most futuristic of the chemical assessment approaches within CompTox.

For additional information from EPA on Virtual Tissues visit:

CompTox Databases

EPA is creating publicly accessible online databases for both new chemical information emerging from the CompTox program and existing information compiled from other sources. These databases include:

  • Aggregated Computational Toxicology Online Resource (ACToR) – ACToR is a repository of all publicly available chemical toxicity data. It also houses information related to chemical risk to human health and the environment. This database contains information on over 500,000 chemicals from more than 650 public sources.
  • Distributed Structure-Searchable Toxicity Database Network (DSSTox) – DSSTox is an online public database of standardized chemical structures that are linked to “high-interest” toxicity data sets.
  • Toxicity Reference Database (ToxRefDB) – ToxRefDB contains data on hundreds of chemicals, mostly pesticides, compiled from thousands of traditional in vivo animal toxicity studies. To date, the studies in ToxRefDB are primarily regulatory “guideline” studies that follow standardized protocols.
  • Toxicity Forecaster Database (ToxCastDB)– The ToxCast database is being built to house data on thousands of chemicals being run through the hundreds of HT screening assays in ToxCast. Only a subset of the ToxCast data is currently available.
  • Exposure-Based Chemical Prioritization Database (ExpoCastDB) – The ExpoCastDB contains measurements of the levels of chemicals in environmental and biological media collected from homes and childcare centers. Data currently available include, to varying degrees, the amounts of these chemicals found in or on food, drinking water, air, dust, indoor surfaces and urine. The public version of ExpoCastDB currently includes limited exposure data for 99 chemicals, mostly pesticides.

For more information on these databases see:

Toxicological Prioritization Index (ToxPi)

ToxPi is a computer software tool that incorporates different sets of toxicity data on a chemical and produces a visual output of those data—a so-called “Pi” chart. ToxPi seeks to convey the “comprehensive toxicity” of a chemical in a manner that can be quickly ascertained through visual representation and be used for prioritizing chemicals. The ToxPi tool is flexible, allowing the user to choose which datasets to include in the construction of a Pi.

ToxPi incorporates multiple data sets (left figure) to produce a visual representation of comprehensive toxicity for a specific chemical (right figure). EPA envisions using “Pis” to help prioritize chemicals. Image Source: Reif et al. (2010). “Endocrine Profiling and Prioritization of Environmental Chemicals Using ToxCast Data.” Environmental Health Perspectives.

For more information on ToxPi visit:

To learn about the possible benefits of these new technologies, proceed to Section 5: Potential of High-Throughput In Vitro Approaches