Accelerate Your Solubility Screening: fastsolv is Now Integrated into Vecura
This update enables medicinal chemists, formulation scientists, and drug discovery teams to predict solid solubility with ease through a guided, no-code workflow inside Vecura, removing the need for complex technical or environment setup.
What is fastsolv?
fastsolv is a powerful, deep-learning-based tool designed to predict the solid solubility of organic small molecules in various solvents at specific temperatures. By leveraging an ensemble of neural networks trained with Sobolev regularization, it produces reliable estimates of logarithmic solubility (logS) alongside calibrated uncertainty metrics. It effectively streamlines drug discovery and formulation science by eliminating the need for computationally expensive 3D conformer generation.
It helps users rapidly estimate solubility, which is critical for evaluating active pharmaceutical ingredients (APIs) and optimizing solvent selection. It is especially useful for researchers looking to screen virtual chemical libraries, rank intermediates, or flag compounds with poor bioavailability early in the development lifecycle.
What can users do with fastsolv on Vecura?
With fastsolv on Vecura, users can:
- Predict the thermodynamic solid solubility of solute-solvent pairs at user-defined temperatures.
- Assess prediction reliability using built-in aleatoric uncertainty estimates (
predicted_logS_stdev). - Integrate solubility screening directly into virtual compound library workflows.
- Reduce reliance on time-consuming wet-lab solubility experiments for early-stage screening.
What the output means
The output provides a predicted_logS value, which represents the base-10 logarithm of molar solubility, and a predicted_logS_stdev, which indicates the uncertainty of the prediction based on ensemble variance.
This output should be used to support scientific decision making. It does not replace experimental validation.
Why this matters
In drug development, predicting solubility is essential for successful API formulation and ensuring the bioavailability of new therapeutic candidates. Traditional experimental methods are often slow and resource-intensive, creating bottlenecks in the development pipeline.
By providing a fast, computationally efficient alternative, fastsolv allows scientists to make data-driven decisions much earlier. This minimizes the need for extensive wet-lab testing and ensures that only the most promising candidates proceed through development, ultimately accelerating the discovery of effective treatments.
- Developed by: Jackson Burns et al.
- Source: GitHub Repository
- Reference: fastsolv Web Application
Vecura で FastSolv を試す。
モデルワークスペースを開き、ご自身の入力で評価を始めましょう。