rowan

Rowan - Cloud-Native Molecular Modeling and Drug Discovery Platform

Overview of Capabilities

Rowan is a cloud-native molecular modeling platform for medicinal chemists and computational chemistry researchers. Through a Python API, it provides end-to-end drug discovery workflows including pKa prediction, molecular docking, conformational search, protein–ligand co-folding, and molecular dynamics simulations.

Use Cases

1. Lead Optimization and Screening

When you need systematic optimization of lead compounds, Rowan offers a complete drug design toolchain. You can run a multi-step workflow of conformational search → molecular docking → pose analysis, handling tens to hundreds of similar compound series—without having to maintain an on-premises GPU cluster or install multiple standalone software packages. All computations run in the cloud, results are stored persistently, and it supports batch screening and SAR analysis.

2. Batch Drug Property Prediction

For scenarios requiring large-scale ADMET assessment, Rowan supports batch job submission workflows. You can compute properties for hundreds of compounds at once—such as descriptors, pKa, permeability, and solubility—with all tasks executed in parallel and results centrally managed. Free users receive 20 credits per week, which is enough for small-scale screening and testing.

3. Molecular Simulations Without Local Infrastructure

When your lab lacks HPC or GPU resources, Rowan lets you run compute-intensive tasks such as quantum chemistry, molecular dynamics, and AI-based structure prediction using a simple Python API. The platform automatically handles infrastructure setup, job scheduling, and result storage, so you can focus on the scientific questions.

Core Features

1. Small-Molecule Modeling and Property Prediction

Rowan provides a comprehensive small-molecule modeling toolkit, including:

pKa prediction: supports microscopic pKa for individual ionizable sites and macroscopic pKa across a range of pH values

Conformational search: generates diverse sets of 3D conformations for docking or dynamics simulations

Tautomer search: handles tautomers in heterocycles and keto–enol systems

Descriptor calculation: 200+ molecular descriptors including MW, LogP, TPSA, Lipinski, and more

ADMET prediction: drug-like properties such as permeability, solubility, and membrane permeability

2. Structure-Based Drug Design

Structure-based drug design workflows:

Molecular docking: supports single-ligand docking and docking for analog series

Protein–ligand co-folding: predicts bound complexes using AI when no crystal structure is available

MD-based pose refinement: refines docking poses with molecular dynamics

Binding free energy perturbations: computes relative binding free energy (ΔΔG)

MSA generation: generates multiple sequence alignments for co-folding

3. Batch Workflows and Automation

Rowan is designed for programmatic batch processing:

Batch submission: submit hundreds of workflows via Python loops

Project organization: manage large screening activities using projects and folders

Webhook notifications: automatically notify your backend when asynchronous workflows complete

Persistent results: all results are stored in the cloud, supporting UUID lookup and reuse

Frequently Asked Questions

What is the Rowan Molecular Modeling Platform?

Rowan is a cloud-native computational chemistry and drug discovery platform that provides a unified molecular modeling interface via a Python API. Unlike traditional software that requires local installation and maintenance (e.g., Schrodinger, AutoDock), Rowan runs all computations in the cloud. With just an API key, you can run complete workflows from descriptor calculation to molecular dynamics. The platform uses a credit-based system: free users get 20 credits per week, and paid users can purchase credits as needed.

How do I get started with Rowan for drug screening?

First install the Python client: uv pip install rowan-python. Then set your API key (environment variable ROWAN_API_KEY or rowan.api_key = "..." in code). The simplest way to start is to submit a descriptor workflow to test the connection, and then select specific workflows as needed (docking, pKa, conformational search, etc.). For large-scale screening, create projects and folders to organize workflows. After batch submission, collect results using result() or stream_result().

What molecular modeling workflows does Rowan support?

Rowan supports 20+ workflows covering the full drug discovery lifecycle: core modeling (descriptors, pKa, MacropKa, conformational search, tautomer search), structure-based design (docking, analog docking, co-folding, MD pose refinement), advanced computational chemistry (quantum chemistry, electronic properties, BDE, redox potential, spin states), reaction chemistry (transition state search, IRC), binding free energies (FEP/perturbation), and sequence analysis (MSA). All workflows follow a unified “submit → wait → retrieve” pattern, and support webhooks and project organization.