PharmSD

PharmSD predicts formulation properties of solid dispersions to guide design for improved in vitro dissolution and oral bioavailability of poorly water‑soluble drugs.

Key Features:

• Machine learning predictive modeling: Employs advanced machine learning algorithms to create predictive models for formulation attributes.
• Attribute predictions: Independently predicts physical stability, dissolution type, and dissolution rate of solid dispersions.
• In vitro dissolution prediction: Predicts in vitro dissolution behavior to assess dissolution performance.
• Virtual screening pipeline: Enables virtual screening of drug‑polymer combinations under different experimental conditions to identify optimal formulations.
• Applicability domain assessment: Evaluates model application domains to determine relevance across scenarios.
• Dissolution curve similarity: Calculates dissolution curve similarity to compare formulation outcomes.

Scientific Applications:

• Formulation design optimization: Guides selection of drug‑polymer combinations to improve dissolution and oral bioavailability of poorly water‑soluble drugs.
• Stability assessment: Predicts physical stability risks for candidate solid dispersion formulations.
• In vitro performance comparison: Compares predicted dissolution profiles across formulations using dissolution curve similarity metrics.
• High‑throughput virtual screening: Enables evaluation of multiple formulation scenarios under varied experimental conditions.

Methodology:

Uses advanced machine learning algorithms to build predictive models and integrates these models to produce independent predictions of physical stability, dissolution type, and dissolution rate; performs virtual screening of drug‑polymer combinations under different experimental conditions; assesses model application domains and computes dissolution curve similarity.