infiniSee xREAL

infiniSee xREAL is a Chemical Space navigation platform developed by BioSolveIT that provides exclusive access to Enamine's xREAL Space — the largest catalog of accessible, drug-like compounds available for early-stage drug discovery. The platform is purpose-built to efficiently screen a trillion-scale combinatorial compound collection, enabling researchers to identify relevant chemistry without the need for supercomputers, server clusters, or GPU arrays. All computations run locally on the user's own hardware, ensuring that confidential data and intellectual property remain fully under the user's control.

infiniSee xREAL is the result of a long-standing collaboration between BioSolveIT, a pioneer in chemical space screening technology, and Enamine, a globally trusted compound supplier. The platform is designed for medicinal chemists, computational chemists, and drug discovery teams working on ligand-based, fragment-based, and structure-based drug design, as well as those integrating AI and machine learning approaches into their workflows.

The xREAL Space

• Contains 4.4 trillion accessible, drug-like compounds in its latest version (infiniSee xREAL 7.0 — Arke Prime), making it several times larger than the original REAL Space.
• Built from Enamine's extensive building blocks and reactions, delivering unparalleled chemical diversity.
• Maintains high synthesis success rates exceeding 80%, ensuring compounds are genuinely accessible.
• Continuously improved using machine learning approaches to enhance predictive accuracy.
• All retrieved compounds are commercially available from Enamine at competitive pricing and can be delivered within a few weeks.
• The Chemical Space file is downloaded directly within the software interface after installation, decoupled from the software download itself.

Hardware and Deployment

• Runs entirely on local hardware — no internet connection, server, or cloud infrastructure required.
• Minimum recommended setup: four CPU cores and 32 GB of RAM; more cores and RAM improve performance due to parallelized algorithms.
• A local graphics card is mandatory for running infiniSee xREAL.
• No data is shared with third parties; all results and queries remain on the user's own system.
• infiniSee xREAL exclusively screens the xREAL Space; other make-on-demand Chemical Spaces can be explored using the standard infiniSee version.

Search Modes and Core Capabilities

• Scaffold Hopper: Retrieves compounds based on fuzzy pharmacophore features, enabling exploration beyond close structural analogs.
• Analog Hunter: Identifies compounds based on molecular fingerprint similarity to a query molecule.
• Motif Matcher: Extracts molecules based on maximum common substructure or an exact substructure defined by the user, supporting fragment-based approaches.
• Multiple orthogonal search methods can be combined to improve enrichment of hit candidates.

Drug Discovery Use Cases

• LBDD (Ligand-Based Drug Design): Retrieve compounds based on molecular similarity using multiple search methods for orthogonal screening.
• FBDD (Fragment-Based Drug Design): Screen for compounds containing a specific fragment as an exact substructure, vastly exceeding the scope of enumerated lists.
• SBDD (Structure-Based Drug Design): Generate chemically diverse subsets of commercially available molecules for ultra-large virtual screening and docking studies.
• SAR Analysis: Discover close analogs with diverse modifications to explore the chemical space around a compound of interest.
• AI/ML Integration: Translate de novo or machine learning-generated suggestions into tangible, accessible compounds by finding closely related structures in xREAL.
• Custom Library Generation: Build tailored compound libraries for specific workflows or AI/ML approaches.

Postprocessing and Analysis (Analyzer Mode)

• Browse and filter search results interactively to identify the most suitable candidates.
• Analyse compound distributions using property histograms.
• Cluster results based on Bemis-Murcko scaffolds and skeletons.
• Export results in multiple formats and create custom libraries of accessible compounds.
• Calculate and filter by a broad range of physicochemical properties including molecular weight, logP, TPSA, H-bond donors and acceptors, heavy atoms, aromatic rings, rotatable bonds, stereo centers, halogens, and more.

ADME Property Prediction (via Optibrium StarDrop Integration)

• CYP2C9 pKi: Predicts affinity of a compound to bind at the CYP2C9 enzyme involved in metabolic drug pathways.
• CYP2D6 affinity category: Classifies compounds into four categories (low, medium, high, very high) based on predicted pKi values.
• BBB category and log([brain]:[blood]): Classifies compounds for blood-brain barrier penetration; useful as an indicator for CNS-active compounds.
• HIA category: Predicts human intestinal absorption, classifying compounds as absorbed ≥30% or <30%.
• P-gp category: Predicts whether a compound is a substrate of P-glycoprotein (MDR1/ABCB1).
• PPB90 category: Classifies plasma protein binding as low (<90% bound) or high (>90% bound).
• hERG pIC50: Predicts inhibition of hERG potassium channels, a key cardiac safety parameter.
• logD: Logarithm of the n-octanol-water partition coefficient at physiological pH 7.4 for ionized compounds.
• logP: Logarithm of the n-octanol-water partition coefficient for neutral compounds.
• logS and logS @ pH 7.4: Predicts intrinsic aqueous solubility for neutral and ionized compounds respectively.

Compounds identified through infiniSee xREAL can be ordered directly from Enamine by submitting structures in SMILES or SD format along with the compound ID and requested amount to [email protected]. The platform is available for download with the Chemical Space accessible directly through the software interface post-installation, and a Beginner's Guide is provided to support new users in getting started quickly.