Computational Toxicology

ILS staff utilize their extensive expertise in computational toxicology, data mining, bioinformatics and cheminformatics approaches to identify and evaluate methods that increase efficiency and allow a more mechanistic understanding of potential adverse effects.

The computational toxicology group at ILS offers a wide range of expertise across informatics areas, allowing us to take data from your ILS assays, in-house resources, and literature to deliver decision-ready products.

ILS has significant experience with a variety of QSAR platforms, using both commercially available and open source software to identify the presence of structural alerts in a chemical under evaluation that are positively or negatively associated with an observed activity (e.g., genotoxicity).

ILS scientists also analyze Tox21 and ToxCast data from quantitative high throughput screening (qHTS) and high content (HC) assays using a variety of existing tools and software, as well as novel computational workflows developed and applied by ILS scientists. These unique workflows can be created to account for continual updating of source databases throughout the course of a study, ensuring flexibility to integrate client feedback. Our expertise enables our clients to make this complex work approachable and understandable through international meetings and peer-reviewed journal articles.

Faced with limited toxicological data for a variety of chemicals present in the environment, quantitative structure activity relationship (QSAR) programs have become more important in prioritizing chemicals and studies.

QSAR is a rapid and inexpensive approach for identifying the potential health hazards of chemicals with limited toxicological data, and can be helpful for prioritizing in vitro and in vivo studies for further testing. ILS brings a wide range of expertise in QSAR modeling and analyses demonstrated by publications that detail QSAR models for estrogen receptor (ER) binding, skin permeability, and skin sensitization (Zang et al. 2013; Braga et al. 2017; Alves et al. 2016; Alves et al. 2015a, b), and high-profile efforts in support of the National Toxicology Program SAR predictions of toxicity of chemicals spilled into the Elk River in West Virginia in 2015. ILS staff also built QSAR models for endpoints such as hepatic clearance and plasma protein binding using published experimental data, physicochemical properties and structural fingerprints, including existing libraries of structural descriptors.

ILS staff are skilled at performing reverse toxicokinetic (RTK) modeling and in vitro to in vivo extrapolation (IVIVE) using metabolic clearance and plasma protein binding data with population-based PK models to quantitatively compare in vitro and in vivo dosimetry for environmental chemicals that potentially interact with pathways of interest, such as the ER and androgen receptor (AR). More detailed PBPK models for specific chemicals, either reference substances or environmental chemicals of concern, can be developed to assist in interpretation and application of in vitro assays that map to specific adverse outcome pathways (AOPs) or may be associated with toxicity endpoints in vivo.

  • Analysis and application of Tox21 and ToxCast data from quantitative high throughput screening (qHTS) and high content (HC) assays
  • Analysis of qPCR, microarray, NextGen sequencing data
  • In vitro to in vivo extrapolation including PBPK modeling
  • Development of integrated approaches to testing and assessment
  • Adverse outcome pathway (AOP)-based testing and data anchoring

  • ReadAcross
  • Quantitative structure property relationship (QSPR)
  • Quantitative structure activity relationship (QSAR)

  • Literature review and background document preparation
  • Expert-led data extraction and curation
  • Database development
  • Data mining and visualization to support safety assessments