infiniSee

infiniSee is a Chemical Space navigation platform developed by BioSolveIT, designed to help drug discovery teams explore ultra-vast collections of synthetically accessible compounds. By leveraging cutting-edge algorithms, infiniSee enables researchers to mine compound collections exceeding billions — and even up to 10¹⁴ structures — delivering results within seconds on standard hardware. It is purpose-built for early-stage small molecule drug discovery, providing purchasable compounds and novel intellectual property with unprecedented speed and ease.

Each search mode in infiniSee retrieves molecules based on the specific needs of the project, allowing scientists to cherry-pick the best candidates. Where conventional approaches may take minutes to hours to process enumerated lists of millions of entries, infiniSee completes searches in typically less than a minute on a standard laptop, making it an indispensable tool for medicinal chemists, computational scientists, and drug discovery teams across academia and industry.

Core Search Modes

• Scaffold Hopper: Screens ultra-large Chemical Spaces for molecules related by fuzzy pharmacophore similarity to a structure of interest, powered by the FTrees algorithm. This mode provides molecules with high novelty and IP potential while preserving the pharmacophore flavor of the query.
• Analog Hunter: Retrieves close analogs to a query compound based on Tanimoto fingerprint similarity using the SpaceLight algorithm. Ideal for exploring the proximal Chemical Space around a compound to elucidate structure-activity relationships (SARs) and identify molecule variants with improved ADME properties.
• Motif Matcher: Mines for compounds containing a particular substructure of interest or those sharing the Maximum Common Substructure (MCS) with a query compound, powered by the SpaceMACS algorithm. Supports both MCS-based searches and exact substructure searches, unlocking opportunities for fragment-based drug discovery (FBDD).
• Analyzer: Manages results and provides deep insight into compound sets. Clusters molecules based on Bemis-Murcko scaffolds and skeletons, visualizes data in parameter histograms (e.g., MW, logP, H-bond donors/acceptors), and filters for drug-like, fragment-like, or lead-like compounds.

Key Capabilities and Differentiators

• Unprecedented speed: Navigate Chemical Spaces of 10¹⁴ molecules within minutes on standard hardware, with no need for high-performance computing infrastructure.
• Visual similarity understanding: Similarity between query and hit molecules is intuitively color-coded, allowing users to immediately understand why infiniSee considers a compound related to the query.
• Scaffold hopping: Discovers molecules that appear distant at first sight yet are very close in chemistry and pharmacophore-based action, enabling the identification of new scaffolds for approved drugs and patented molecules.
• Accessible results: Only compounds highly likely to be synthesized in the lab are proposed. Hits from partner Chemical Spaces are purchasable and can be delivered within a few weeks.
• Firewall-safe operation: infiniSee runs on almost any computer with modest hardware requirements, operating securely behind the user's firewall.
• Fine-tuned searches: Users can apply different parameters, constraints, and fuzzy pharmacophore definitions to tailor searches and extract the most relevant compounds.
• Combinatorial search approach: Instead of searching pre-assembled molecules, infiniSee performs a combinatorial build-up of compounds from fragments, enabling the screening of billions of compounds through its similarity search technology.
• Multi-space searching: Searches can span multiple Chemical Spaces of different origins, significantly increasing the chances of finding accessible molecules that would otherwise be missed.

Analyzer Mode: Compound Clustering and Data Exploration

• Cluster compounds based on Bemis-Murcko scaffolds and skeletons to select chemically diverse candidates.
• Explore parameter histograms for properties such as molecular weight, logP, and number of H-bond donors and acceptors.
• Filter directly for drug-like, fragment-like, or lead-like compounds.
• Investigate reagents and building blocks associated with each compound.
• Group compounds based on their vendors for streamlined procurement decisions.

Supported Chemical Spaces

• REAL Space by Enamine
• AMBrosia by Ambinter
• CHEMriya by OTAVAchemicals
• eXplore by eMolecules
• Freedom Space by Chemspace
• GalaXi by WuXi LabNetwork
• KnowledgeSpace — a literature-based virtual space with high likelihood of synthetic accessibility, offered by BioSolveIT for ideation purposes.

Creating Custom Chemical Spaces

• Users can design their own Chemical Space using in-house reactions and building blocks to create massive numbers of virtual, synthetically accessible compounds.
• CoLibri, BioSolveIT's companion toolkit, transforms synthetic know-how into a Chemical Space, enabling pharmaceutical teams to encode their proprietary chemistry knowledge.
• Multiple large pharmaceutical companies have built in-house Chemical Spaces using this approach to search for scaffold alternatives, reducing costs and development time.
• BioSolveIT also offers a service to create bespoke Chemical Spaces tailored to a client's proprietary building blocks and chemistry knowledge.

infiniSee version 7.0 — codenamed Arke — introduces a new database format for ultra-large Chemical Spaces, enabling even more efficient screening. The platform integrates seamlessly with BioSolveIT's broader ecosystem, including SeeSAR for drug design and HPSee for high-performance computing workflows, and supports partnerships with multiple compound manufacturers to ensure broad chemistry coverage and flexible compound procurement.