bgbg
beta

Molecular simulation toolkit by BiosimAI

Physicochemical simulations and AI for interrogating biology at the atomistic scale.

Develop safe, effective drugs faster.

Toolkit for in silico drug discovery

Toolkit for in silico drug discovery

The BiosimAI toolkit enables scientists to interrogate biology from individual atoms to whole cells using biological simulations, biophysics, and machine learning. With our toolkit, scientists can develop safe, effective drugs faster.

Our toolkit includes:

  • Atomistic ligand docking
  • Molecular properties AI: solubility, toxicity, etc.
  • Atomistic molecular dynamics

The toolkit is available through Jupyter notebooks in BiosimAI software blueprints on the Deep Origin cloud platform.

Ligand docking and molecular properties

Docking places candidate molecules within a protein's binding pocket and estimates their binding affinities. By integrating physics and AI, our algorithms offer state-of-the-art accuracy and speed.

  • Dock compounds against multiple protein conformations.
  • Identify stable poses of compounds in binding pockets.
  • Estimate protein-ligand binding affinities.
  • Predict properties such as solubility, toxicity, and more.
Ligand docking and molecular properties
The R&D platform for scientists, by scientists

Atomistic molecular dynamics

Molecular Dynamics (MD) simulations enable the study of multi-protein complexes, the discovery of hidden binding sites, and precise estimation of binding affinities between compounds and proteins.

Our tools combine physics simulation and AI, enabling:

  • Simulating proteins, small organic molecules, and protein-drug complexes.
  • Explicit solvent simulations using the AMBER and CHARMM force fields for accurately predicting protein folds.
  • AI-approximated quantum chemistry force field for accurately modeling small molecules.
  • State-of-the-art free energy perturbation (FEP) methods for accurately estimating binding affinities.
  • Enhanced sampling methods, such as replica and umbrella sampling, for efficiently estimating binding affinities.
icon

Scientist-Friendly Interface

  • Orchestrate simulations with Julia in JupyterLab and VS Code.
  • Get started in minutes by running our simulation toolkit in our computational platform.
icion

Protein binding

How compounds bind proteins and how strongly they bind

icon

Solubility & distribution

Ability of compounds to circulate to target tissues

icon

Toxicity

Adverse, off-target effects of compounds

icon

Protein structure

3D configurations of polypeptides and complexes

icon

Biological simulation

Phenomenological simulation of biological entities and their interactions

icon

Physics simulation

Ab initio simulation of chemical and mechanical forces

icon

Machine learning

Data-driven simulation of behaviors, from individual molecules to cells

icon

Whole-cell simulation

Predict phenotypes of pathways to whole cells

icon

JupyterLab

Web IDE for notebooks, code, and data

icon

VS Code Server

Extensible web IDE for data analysis and coding

icon

Julia

High-level, dynamic language

icon

Python

High-level, general-purpose language

ENTERPRISE AND PROFESSIONAL SERVICES

High-throughput virtual screening

With our docking, ADMET, and molecular dynamics, we can help you identify lead compounds by cost-effectively screening billions of molecules. Contact us to learn more.

bgbg