(General Unrestricted Structure-Activity Relationships) web service on the Way2Drug platform is designed for quantitative prediction of biological activity and acute toxicity of organic compounds. The online tool is based on a computational methodology that uses QNA structural descriptors (Quantitative Neighborhoods of Atoms) and PASS bioactivity profiles as independent variables to build high-accuracy consensus models using the Self-Consistent Regression (SCR) method. GUSAR has been developed according to OECD principals and includes last achievements in the field of QSAR modeling: consensus prediction, applicability domain assessment, internal and external models validation and clearly interpretations of obtaining results.
For user convenience, prediction in the web service is divided into three specialized sections:
Acute Rat Toxicity Prediction. This module estimates the median lethal dose (LD50) in rodents for four different routes of administration: oral, intravenous, intraperitoneal, and subcutaneous.
Prediction of Antitargets Interaction Profiles. This section estimates the likelihood that the tested molecule binds to specific proteins—receptors, enzymes, and transporters. Such interactions with antitargets can lead to serious adverse effects in humans.
Ecotoxicity Prediction. This tool is intended to assess the potential negative environmental impact of chemical compounds. It supports predicting toxicity for aquatic organisms and estimating bioaccumulation potential.
After uploading the molecular structure data in standard SD file format, the algorithm automatically calculates descriptors and generates a prediction. The prediction results are presented to the user in the form of a detailed table, as shown in the provided screenshots. This table displays the predicted numerical values for specific biological activities (e.g., Log10(Value) for antitargets or Log10(mmol/L) for environmental toxicity) alongside an "Applicability Domain" (AD) assessment. The AD column explicitly indicates whether the tested compound falls within the reliable predictive scope of the model ("In AD" highlighted in green) or outside of it ("Out of AD" highlighted in red). This immediate AD evaluation helps researchers determine the statistical reliability of the prediction for each specific target or endpoint before making decisions about compound safety.
GUSAR tools are actively used at early stages of drug discovery for large-scale virtual screening of chemical compound databases. In silico analysis allows researchers to filter out obviously toxic or environmentally hazardous molecules before they are synthesized in the laboratory. This approach significantly reduces costs, accelerates the development of safer compounds, and can substantially decrease the number of experiments performed on laboratory animals.
Quickly estimate risks (acute toxicity, ecotoxicity, and potential off-target/antitarget interactions) for new or virtual compounds before you spend time and money on synthesis and wet-lab testing.
It provides quantitative predictions (not just toxic / non-toxic), so you can rank compounds, prioritize the safest candidates, and compare alternatives within a project.
Improving workflow for early-stage screening by enabling fast filtering of large libraries and highlighting liabilities (e.g., antitarget binding signals) early, which reduces late-stage failures.
Alexey Lagunin et al. (2011)
QSAR Modelling of Rat Acute Toxicity on the Basis of PASS Prediction.
Molecular Informatics, 30(2-3):241-50.
If you need to evaluate a large dataset, or if you need to maintain confidentiality of structural formulas transmitted via unsecured data channels, you can contact us to discuss the licensing opportunities.
© 2011 - 2026 Way2Drug. All rights reserved.