Based on multimodal 3D generation, seeking novel, structurally sound molecules that fit target pockets dynamic optimization of candidate molecules, pursuing more efficient human-machine interaction
Molecule Generation
MD-IFP
RBFE
Torsion Scan
Substructure Search
Multimodal AI 3D molecule Generation Model
The molecule generation model of MolVado™ enables empty pocket-based or substructure-constrained 3D molecule design. It can deliver novel and chemically valid compounds. Through considered structure and binding rationality, researchers can efficiently optimize molecules from pocket to drug candidates.
De Novo generation based on empty pocket
Generation based on fixed fragments of reference molecules
Interaction Fingerprint Analysis of Molecular Dynamics Trajectories (MD-IFP Analysis)
The AI-powered MD-IFP analysis tool of MolVado™ overcame the limitations of static structures.
By utilizing molecular dynamics simulations with AI-assisted analysis, the tool enables researchers to reveal key conformational changes and automatically identify the representative binding poses.
Relative Binding Free Energy (RBFE)
The relative binding free energy (RBFE) solution of MolVado™, powered by Thermodynamic Integration (TI), delivered chemically accurate (1 kcal/mol level) predictions of binding free energy differences between structurally similar molecules.
The RBFE calculation tool has been validated across diverse targets, including kinases, apoptosis proteins, proteases, etc., which can provide reliable activity predictions to complement or partially replace experimental measurements.
Torsional Energy Scan
The torsion energy scanning tool of MolVado™ integrates double-hybrid density functional theory (DFT) methods with AI-powered prediction to deliver reliable conformational analysis.
With its proprietary fragmentation algorithm and asynchronous scanning algorithm, it can minimize the impact of structural errors on torsional energy scanning results, reliably explore low-energy conformations, and effectively capture the most stable conformational states. In addition, it natively supports implicit solvent models, enabling direct simulation of biologically relevant environments, bridging the accuracy gap between gas-phase calculations and condensed-phase experimental observations.
Substructure Search
MolVado™ has a comprehensive database of more than 10 million drug-like small molecules from global patents and literature with an over 95% coverage rate and monthly updates.
Users can select molecular fragments generated by the model or upload themselves and directly link to key information, such as patent numbers, assignees, publication details, and literature DOIs.
Swirling historical data into breakthrough innovation
Customized data-driven platform for drug R&D requirements
MolVortex® builds on the MolVado™ platform to deliver customized AI solutions. For proprietary data curation and usage without exposing, MolVortex® transforms historical project data into the vector database and integrates the large language model (LLM) to retrieve the acknowledgment for subsequent generation.
Scalable to a wider range of R&D modalities
In addition, based on different data, MolVortex® has been developed as a new generation model for other modalities, such as covalent compounds, peptides, molecular glues, PROTACs, etc.