There are over 350,000 chemicals registered for commercial use globally, but only a fraction have been thoroughly tested for environmental safety. Traditional testing — exposing organisms to chemicals in laboratories — is slow, expensive, and raises ethical concerns. In silico testing offers a transformative alternative.
What Is In Silico Testing?
"In silico" means "in silicon" — a reference to computer chips. In silico testing uses computational models to predict how chemicals will behave in the environment and interact with biological systems. It's the digital counterpart to in vitro (test tube) and in vivo (living organism) testing.
How It Works
In silico approaches include:
- **QSAR models** (Quantitative Structure-Activity Relationships) — predicting a chemical's toxicity based on its molecular structure
- **Molecular docking** — simulating how a chemical binds to biological receptors
- **Physiologically-based pharmacokinetic (PBPK) models** — predicting how chemicals distribute through an organism's body
- **Ecosystem models** — simulating how chemicals move through food webs and environments
These models draw on vast databases of existing experimental data, using machine learning and AI to identify patterns and make predictions for untested chemicals.
Advantages
Speed is the biggest advantage. Testing a single chemical in the laboratory can take years and cost millions. In silico screening can evaluate thousands of chemicals in days. This is crucial given the enormous number of chemicals in commerce that haven't been assessed.
The approach also reduces animal testing — an increasingly important ethical and regulatory consideration. The EU's REACH regulation actively encourages the use of computational methods.
The CONTRAST Contribution
CONTRAST researchers are developing in silico tools specifically for predicting the effects of CECs on marine organisms. These models account for the unique challenges of the marine environment — species that don't have standard test protocols, complex exposure scenarios, and mixture effects.
Limitations and the Future
In silico models are powerful but not perfect. They require high-quality data to train on, and novel chemical structures may fall outside their predictive domain. The future lies in integrating in silico with in vitro and targeted in vivo testing — a "tiered" approach that maximises information while minimising cost and animal use.
This article is part of the CONTRAST project, funded by the European Union under Horizon Europe. Views expressed are those of the author(s) only.