Computational Chemistry
Schrödinger
by Schrödinger, Inc.
Physics-based molecular simulation platform accelerating drug discovery and materials design
Category
Computational Chemistry
Founded
1990
Headquarters
New York, NY, USA
Overview
Schrödinger provides industry-leading computational chemistry software combining physics-based molecular simulations with machine learning to predict molecular properties, optimize drug candidates, and design novel materials. The platform includes FEP+ (free energy perturbation), Glide (molecular docking), Maestro (molecular modeling interface), and LiveDesign (collaborative medicinal chemistry). Pharmaceutical companies use Schrödinger's platform to prioritize compounds before synthesis, reducing experimental cycles by 50% or more. The platform is used by 19 of the top 20 pharmaceutical companies and has contributed to multiple clinical-stage drug programs. Schrödinger also operates an internal drug discovery pipeline leveraging its own technology. The combination of rigorous physics-based modeling with modern ML approaches gives Schrödinger a unique position. FEP+ predictions routinely achieve sub-1 kcal/mol accuracy for binding free energy calculations. The company's dual model — software licensing plus internal drug discovery — demonstrates confidence in its own computational platform.
Key Features
Reaction Pathway Analysis
Computational prediction of chemical reaction mechanisms and transition state geometries.
Cloud-Scalable Architecture
Elastic cloud infrastructure handles enterprise-scale virtual screening with on-demand scaling.
Molecular Dynamics Simulation
Physics-based molecular simulations predict binding affinities and conformational dynamics.
Free Energy Perturbation
Accurate rank-ordering of drug candidates using free energy calculations with GPU acceleration.
Quantum Mechanics Engine
Quantum mechanical calculations for accurate electronic structure and reactivity predictions.
Pros & Cons
Pros
- +Free energy perturbation calculations accurately rank-order drug candidates reducing wet-lab testing
- +Cloud-scalable architecture handles enterprise-scale virtual screening campaigns
- +Validated against thousands of experimental datasets ensuring prediction reliability
- +Supports multi-target drug design workflows from hit identification through lead optimization
- +Physics-based molecular simulations predict binding affinities with near-experimental accuracy
- +GPU-accelerated calculations enable screening of millions of compounds in hours instead of weeks
Cons
- −Integration with existing drug discovery workflows requires custom pipeline development
- −Requires significant computational infrastructure (GPU clusters) for large-scale molecular simulations
- −Steep learning curve demands expertise in both computational methods and medicinal chemistry
- −Prediction accuracy varies significantly across different protein targets and binding site types
- −Enterprise licensing costs can exceed $100K/year making it inaccessible for academic groups
Use Cases
Molecular Dynamics Simulation
Physics-based molecular simulations predicting binding affinities and conformational dynamics with near-experimental accuracy.
Free Energy Perturbation
Accurate rank-ordering of drug candidates using free energy calculations to reduce wet-lab testing requirements.
Virtual Compound Library Design
GPU-accelerated screening of millions of compounds to design focused libraries for synthesis and testing.