Molsoft

ICM-VLS ( Virtual Ligand Screening) is MolSoft’s efficient virtual screening methodology in which ligands are fully and continuously flexible, some of the core functionalities include:

1. Ligand Docking – Batch
2. Covalent Docking
3. Structure-based VLS
4. Virtual Ligand Screening Physics/AI based scores
5. Ligand-based VLS (APF)
6. Fragment-based VLS
7. Dock on the fly Markush generated libraries
8. ICM 3D Interactive Editor
9. Atomic Property Fields (APF)
10. RIDE (CPU version)
11. Pharmacophore Searching
12. Includes all Functionalities of ICM Chemist – Pro

To know more click here.

ICM-Homology offers a methodology for conformational modeling of protein side chains and loops, implemented in ICM, relies on internal coordinate definition of the molecular object combined with computationally efficient ICM Biased Probability Monte Carlo

1. Sensitive sequence search for template identification.
2. Fast model building – the algorithm builds the model with all loops in seconds.
3. Loop prediction using a loop PDB database.
4. Loop prediction through local energy optimization.
5. Multi-template modeling.
6. Loop grafting.
7. Protein sculpting.
8. Peptide modeling.
9. Search the PDB for similar loop conformation.
10. Calculate relative residue frequency in similar loops from the PDB.
11. Prediction of Disulfide bonds.
12. Local reliability prediction and model validation features.
13. Model refinement using ICM global optimization.
14. Membrane protein modeling.
15. High-throughput homology modeling.

To know more click here.

Below are concise feature bullets for RIDGE (Rapid Docking GPU Engine):

1. GPU-accelerated structure-based docking
2. Ultra-fast virtual ligand screening (~100 compounds/sec on RTX GPUs)
3. ~1000× speed of a single CPU core
4. Screens 10 Million compounds on a desktop GPU in a day
5. Supports multiple GPUs for parallel jobs
6. Compatible with existing ICM docking projects
7. Uses highly-compressed conformer libraries
8. Neural-network scoring (RTCNN)
9. Option to apply physics-based docking scores
10. Minimal setup from standard ICM dock projects
11. Efficient for ultra-large library screening
12. Works on Linux and Windows platforms

To know more click here.

The GigaScreen method combines machine learning and deep learning tools to tackle the computational intensity of screening very large chemical databases. Key features include:

1. Ultra-large library screening framework
2. Deep learning-guided hit prioritization
3. Combines ML models with GPU docking
4. Uses RIDGE GPU acceleration for speed
5. Screens billions of virtual compounds
6. Iterative model training and scoring
7. Reduces compute costs and data volume
8. Adjustable score thresholds for filtering
9. Automated batch processing workflows
10. Scales on single workstation or GPU cluster
11. Hybrid physics + AI scoring strategies
12. Active learning loop for model refinement
13. Efficient candidate selection from giga-sized libraries
14. Supports custom filtering rules and constraints
15. Integrates with existing ICM workflows
16. Includes trillions of pre-prepared libraries from multiple vendors

To know more click here.

CombiRIDGE is a new innovative GPU-accelerated solution for high-throughput ligand docking and screening which leverages MolSoft’s generative neural network conformer enumeration method GINGER, ultra-fast GPU docking technology RIDGE and advanced graph neural network scoring ( RTCNN ). The approach allows you to optimize specific R-groups and screen vast ultra-large or combinatorial libraries efficiently.

1. Reaction-based combinatorial library generation
2. Synthesizable compound design
3. Applies known chemical transformation rules
4. Generates diverse, focused libraries
5. Preserves synthetic feasibility
6. Customizable reaction templates
7. Flexible scaffold and building block selection
8. Supports enumerations with stereochemistry
9. Integrates with ICM design workflows
10. Enables rapid ideation of analog series
11. Automated combinatorial enumeration
12. Structural filtering and property constraints
13. Facilitates hit expansion strategies
14. High-throughput library creation
15. Supports custom compound rules
16. Enhances medicinal chemistry productivity
17. Efficiently explores chemical neighborhoods

To know more click here.

V-SYNTHES is a hierarchical synthon-based virtual screening platform that enables efficient exploration of ultra-large, synthetically accessible chemical libraries using flexible ICM docking.

1. Hierarchical synthon-based virtual screening strategy
2. Efficient navigation of ultra-large make-on-demand libraries
3. Screens tens of billions of compounds
4. Uses ICM docking for flexible, accurate scoring
5. Fragment-based initial screening and enumeration
6. Minimal Enumeration Library (MEL) generation
7. Iterative docking and hit refinement
8. 4D flexible-receptor representation support
9. Accelerated docking via hierarchical protocol
10. Applicable to gigascale chemical spaces
11. Prioritizes synthetically accessible candidates
12. Scales on desktop or cluster environments
13. Integrates with ICM-Pro and VLS workflows
14. Supports complex target classes like GPCRs and kinases

To know more click here

RIDE is a fast 3D molecular similarity search method based on Atomic Property Fields. RIDE searches databases of compound conformers for molecules that are isosteric to the query

1. Can screen 1.5Million conformers/sec/GPU
2. GPU-Accelerated 3D Similarity Search
3. Bioisosteric Replacement Engine
4. Rapid Lead Optimization Assistance
5. Scaffold Hopping at Scale
6. Flexible Conformer Matching
7. High-Throughput Virtual Replacement
8. Structure-Aware Chemical Space Exploration
9. Integrates with ICM Design Workflows
10. Supports Large Compound Libraries
11. Fast Performance via Parallel GPU Compute

To know more click here.

1. GPU-accelerated conformer generation
2. Neural network-based internal coordinate model
3. Rapid production of diverse, low-energy 3D conformers
4. Highly scalable for very large libraries
5. Generates high-quality conformers with energy refinement
6. Efficiently handles millions of compounds per day
7. Produces smaller, high-accuracy conformer sets
8. Ideal for 3D workflows (screening, docking, similarity)
9. Integrates with ICM-Pro, RIDE, and RIDGE
   Memory-efficient and robust conformer processing

To know more click here.

Ligand AIDE is a de novo ligand generation workflow based on Artificial Intelligent Design Evolution. It uses an iterative evolutionary strategy to grow and optimize ligands directly in the context of a protein binding site.

1. AI-driven de novo ligand generation workflow
2. Iterative evolutionary design directly in a protein binding site
3. Begins from docked fragment population and grows ligands
4. Uses groupGen neural network for R-group addition and replacement
5. Atom-level substitution via neural atom predictor
6. Re-docks new designs each generation for optimized fit
7. Multi-objective selection using AI score, docking score, and drug-like filters
8. Synthesizability criteria enforce practical chemical designs
9. Typically uses multiple evolutionary cycles (e.g., 3-5)
10. Efficient exploration of chemical space with binding quality focus
11. Integrates with docking workflows for automated design refinement

To know more click here.

MolScreen is a set of high quality 2D fingerprint and 3D pharmacophore models for a broad range of pharmacology and toxicology targets. 

1. Panel of >2,500 high-quality 2D & 3D AI/ML models for pharmacology & toxicity targets
2. Models cover ~1,200 distinct targets
3. Includes 2D fingerprint ML classifiers & regressors
4. 3D Atomic Property Field (APF) pharmacophore models
5. ADMET prediction models (e.g., CYP, hERG, half-life)
6. Supports lead identification & profiling workflows
7. Useful for target identification & drug repurposing
8. Fast scoring with validated performance metrics (e.g., high AUCs for key model types)
9. Works with ICM-Pro + VLS screening automation
10. Enables multi-target vs. multi-compound profiling
11. Models can be run directly or via batch jobs in ICM GUI/CLI
12. Includes classification & regression scoring outputs
13. Adaptable for lead optimization and SAR analysis

To know more click here.

RIDE is a fast 3D molecular similarity search method based on Atomic Property Fields. RIDE searches databases of compound conformers for molecules that are isosteric to the query

1. Can screen 1.5Million conformers/sec/GPU
2. GPU-Accelerated 3D Similarity Search
3. Bioisosteric Replacement Engine
4. Rapid Lead Optimization Assistance
5. Scaffold Hopping at Scale
6. Flexible Conformer Matching
7. High-Throughput Virtual Replacement
8. Structure-Aware Chemical Space Exploration
9. Integrates with ICM Design Workflows
10. Supports Large Compound Libraries
11. Fast Performance via Parallel GPU Compute

To know more click here.