Frequently Asked Questions

ForceADME provides high-quality predictions with accuracy rates averaging between 80-95% depending on the specific property being predicted. Our models are continuously trained and validated on large, diverse datasets of experimental data.

Each prediction comes with interpretations and methodology details that explain how the property was calculated. While these computational predictions are valuable for early-stage drug discovery, we always recommend experimental validation for critical decision-making points in your research.

The Boiled Egg plot is a predictive model that helps visualize and predict a compound's absorption in the gastrointestinal tract and its ability to cross the blood-brain barrier. The plot uses two physicochemical parameters: lipophilicity (WLogP) and topological polar surface area (TPSA).

On the plot, the white region (egg white) represents compounds with high probability of passive intestinal absorption, while the yellow region (yolk) represents compounds with high probability of brain penetration. This visualization helps identify compounds with desirable pharmacokinetic profiles early in the drug discovery process.

GeinPredictor uses advanced machine learning models trained on large bioactivity datasets to predict the interaction between compounds and specific biological targets. The tool analyzes molecular fingerprints and structural features to estimate binding affinity and potential bioactivity.

Our models incorporate data from ChEMBL, PubChem, and other public databases, as well as proprietary datasets. Predictions are provided as probability scores or estimated IC50/EC50 values depending on the target type and available training data.

Our AI Detection tool helps you evaluate whether compounds are likely to have been generated by AI models. This is particularly important as AI-generated molecules may look promising computationally but could have synthetic feasibility issues or other problems when brought to the lab.

The tool analyzes structural features and patterns that are characteristic of AI-generated molecules, providing a probability score and synthetic feasibility assessment. This helps researchers prioritize compounds that are not only promising in terms of predicted properties but also practical to synthesize.

Usage is generally calculated based on the number of molecules processed or analyses performed. Each tool may have specific usage metrics:

• ForceADME: Each molecule analyzed counts as one usage credit
• GeinPredictor: Each bioactivity prediction against a target counts as one usage credit
• AI Detection: Each molecule evaluated counts as one usage credit

Usage limits reset monthly for subscription plans, and you can always view your current usage in your account dashboard.

Yes, we offer significant discounts for academic institutions, universities, and non-profit research organizations. Our academic plans provide all the features of our Professional plan at a reduced rate.

To qualify for academic pricing, you'll need to register with an institutional email address and may need to provide proof of your academic affiliation. Please contact our sales team for more information on academic pricing.

We store the molecular structures you submit to provide you with historical analysis access and to improve our services. This data is associated with your account and is not shared with other users.

If you're concerned about confidentiality, our Enterprise plans include options for data deletion after processing and signed confidentiality agreements. You can also manually delete any analysis from your history at any time.

You can request account deletion at any time from your account settings page. When you delete your account, we will:

• Remove all personal information associated with your account
• Delete all molecular structures and analysis results
• Remove your account from our user database

In accordance with data retention regulations, we may keep certain information for a limited time in our backup systems, but this data is marked for deletion and is not used or accessed.