Fast Substructure Search in Combinatorial Library Spaces

Abstract

We present an efficient algorithm for substructure search in combinatorial libraries defined by synthons, i.e., substructures with connection points. Our method improves on existing approaches by introducing powerful heuristics and fast fingerprint screening to quickly eliminate branches of nonmatching combinations of synthons. With this, we achieve typical response times of a few seconds on a standard desktop computer for searches in large combinatorial libraries like the Enamine REAL Space. We published the Java source as part of the OpenChemLib under the BSD license, and we implemented tools to enable substructure search in custom combinatorial libraries.

RDCanon: A Python Package for Canonicalizing the Order of Tokens in SMARTS Queries

“…Contemporary virtual combinatorial libraries contain billions of synthetically accessible compounds, , and efficient algorithms are necessary to analyze such large chemical spaces in a tractable manner . While the subgraph isomorphism problem has been well studied in both graph and chemical contexts, − it is still the rate-limiting step in chemical search. Techniques such as indexing or fingerprinting can rapidly filter databases to minimize the number of expensive substructure searches to be run but rely on the precomputation of a mask for each compound in the database–a bit set to 0 or 1 depending on the presence of a specific feature or structural pattern .…”

Shape-Aware Synthon Search (SASS) for Virtual Screening of Synthon-Based Chemical Spaces

“…To navigate such unenumerable libraries, one approach involves using generative models to propose in-library molecules based on learned distributions from a subset of the library. , In another approach, similar to how the libraries are defined by the synthons, virtual screening methods evaluate the synthons instead of the full libraries and only instantiate and evaluate products formed from the best synthons, thereby limiting the computational resources required to evaluate such large chemical spaces. Several two-dimensional (2D) molecular graph-based virtual screening methods that rely on evaluation of synthons have been reported to search ultralarge chemical spaces efficiently. − …”

Shape-Aware Synthon Search (SASS) for virtual screening of synthon-based chemical spaces

“…Several 2D molecular graph-based virtual screening methods that rely on evaluation of synthons have been reported to search ultra-large chemical spaces efficiently. [26][27][28][29][30] Only recently, however, has such synthon-based approach been adopted for 3D virtual screening. Two such methods are Chemical Space Docking [14] and V-SYNTHES [9] , which first dock individual synthons to identify promising synthons and then instantiate and evaluate only products formed from those synthons.…”