ChemAxon’s solution for handling Markush structures

Markush structures, or generic structures, are widely used in combinatorial library and chemical patents to define large chemical spaces and to protect intellectual property. ChemAxon provides powerful Markush structure analysis tools, including importing and rapid structure search in patent Markush structure space. As a result of a collaboration with Thomson Reuters, ChemAxon’s technology is 100% compatible with all patent features of the Thomson Reuters’ standards, including the Markush database (MMS), patent document (DWPI) and exemplified structures (DCR) content. All Markush analysis features are available as an add-on for ChemAxon’s JChem technology (JChem Base, JChem Cartridge and Instant JChem).

Markush structure sketching

Markush structures can be drawn using MarvinSketch. It supports Markush structure variations, such as atom list, bond list, position variation bond, link nodes and repetition units. It also supports complex Markush features, such as multiply-connected R-groups, heavily nested R-groups, and homology groups. Numbered attachment points can be added to multiply-connected R-groups to accurately align to the core or their parent R-groups.

Markush Structure Sketching

Markush structure generation

Besides drawing it manually, a Markush structure can also be generated automatically from a molecule library using R-group decomposition. Starting with the definition of the common scaffold pattern, this process identifies all possible R-group substituents that occur in the library. Using these substituents an R-table can be created and saved, or the scaffold can be saved in a Markush structure along with the R-group definitions. R-group decomposition is available as a command line tool within Fragmenter as well as in Instant JChem and as a KNIME node.

Markush Structure Generation

Markush Viewer

For a complex Markush structure with many R-groups, visualization can be challenging. The Markush Viewer can import and display a Markush structure graphically. The Markush core and all R-groups are listed in a hierarchical format. Users are able to click and review different substituents of each R-group.

Markush Viewer

ChemAxon’s Markush Search

Searching in Markush spaces

Markush Search is a powerful tool for structure search in a large chemical space defined by a generic Markush structure, without actually enumerating the explicit compounds. This approach enables the exploration of extremely large Markush spaces, which would be unthinkable by traditional search techiques. In case of multiple substructure queries against a library of Markush structures, an “Overlap Analysis” can be applied, returning the hits for each query structure spearately. ChemAxon’s Markush technology now allows users to submit Markush structures of limited complexity as a query. Query Markush structures can contain the same Markush features that can be present in general Markush structures, which makes the comparison of simple Markush spaces possible.

Markush Searching

Rich visualization features

Hit coloring and automatic structure alignment make the visualization of the search results easier. The features for better visualization include 2D-cleaning for the hit structures. Users may also choose to reduce visualization on the Markush core only; alternatively, non-relevant R-groups can be filtered out to display only essential information or all R-groups can be shown.

Markush relevant R-groups

Markush reduction

For an even “cleaner” display, Markush Reduction can be applied. This feature can add the R-groups matching the query to the original core, so the expanded core contains the substructure query. The R-groups in the expanded core are kept in display while the others are removed. The part of the extended core matching the query is also highlighted.

Markush reduction

ChemAxon Markush Enumeration

Markush Enumeration is a powerful tool for Markush structure analysis and is fully implemented in JChem. By interactively guiding the enumeration of the Markush Structure users can quickly understand the relevance of Markush to their needs. Current enumeration types include; full, partial, and random enumeration; library size calculation; homology enumeration.

Markush Enumeration

Filter enumerated structures

To reduce the often large library sizes after Markush enumeration, the resultant structures can be filtered by predicted properties, using ChemAxon’s many structure based calculations (either singly or combined) to identify structures with suitable chemical properties.

Markush usage Instant JChem

Markush code generation

In the combinatorial chemical application of the Markush technology the unique identification of the enumerated structures can be essential. Therefore, a special ID number can be generated for the library members. Every structure gets its own unique tag (molecule property), which can be saved in the structure file (in .mrv and .sdf formats) named as ‘Markush Code’. This ID is available in the plugin result window.  It can provide structural information, such as R-group number; link node on an atom; bond between atoms; position variation bond between atoms, etc. It may also contain customized identifier and names for the Markush structure.

Markush Code Generation

Thomson Reuters patent database

Markush usage Instant JChem Thomson Reuters and ChemAxon have formed a strategic partnership enabling users for the first time to host, search and analyze Thomson Reuters full IP Data Feeds, including Markush structures. The arrival of this search and analysis solution will speed drug discovery and allow life science researchers and IP experts to easily integrate critical content into their existing systems and workflow. Thomson Reuters Markush DARC (.vmn) files contain Markush structures extracted from patents. Each VMN file contains one Markush structure. These VMN files can be imported to the JChem technology  as Markush structures.

ChemAxon offers assistance with the loading of raw files provided by Thomson Reuters into a cross-linked relational database structure. This database includes textual, Markush and specific (e.g. exemplified) chemical structure data. Instant JChem is an ideal tool to search such a complex database scheme.

Markush structures in a workflow

Markush Code GenerationWorkflow management solutions play a more and more important role in everyday research processes. ChemAxon’s Markush analysis plug-in is implemented in KNIME and in Pipeline Pilot, two commonly used workfow management toolkits. These implementations includes Markush Search, Markush Enumeration, and Markush structure generation which makes this technology easily accessible and integratable into best practice routines.

Articles in the library

Tools for Analyzing Exemplified and Markush Structures in Chemical Patents Presentation

Medicinal chemists rely on patent intelligence at three distinct junctures of a research project: inception of an idea, ongoing competitive intelligence during project development, and pr…
September 2011 Author: ()

Advanced searching chemical structures in patents Presentation

Patents allow you a limited monopoly on your invention. Equally important, they allow your competitors the same monopoly on their inventions. Chemical structures in patents are frequently…
September 2011 Author: ()

Latest Developments in Markush Representation, Search, Analysis and Visu… Poster

Scientific poster at the International Patent Information Conference and Exposition (IPI-ConfEx) Markush structures are widely used in patent claims to maximize intellectual property pro…

Related forum threads

Frequently Asked Questions

What is Markush search good for?
It allows efficient substructure and full structure searching in Markush libraries, and helps visualizing substructure hits.
Who uses Markush search?
Medicinal chemists, patent professionals, or synthetic chemists involved in combinatorial chemistry.
Do I have to do programming for using Markush Search?
Not necessarily. For customization, you may have to do programming.
What kind of additional software is needed?
Java Virtual Machine version 5 or later. A relational database engine (RDBMS) that will store the structures.
Can Markush Search store reactions/polymers/query structures/Markush structures, etc?
No, currently Markush tables cannot hold reactions or polymers. For the list of currently available features, see the JChem Query Guide.
What Markush features are supported?
Markush search currently can handle the following generic features:
  • R-groups (nested to any depth)
  • atom and bond lists
  • link nodes and repeating units
  • position variation bonds
  • homology variation (predefined generic atoms – alkyl, aryl, etc. and user-defined ones, like protecting group)
Can Markush Search handle nested R-groups?
Yes, arbitrary deep nesting is possible, but the references cannot be circular (E.g. R1 definition cannot use R1).
Patents often contain conditions as well. Do you plan to handle these?
Conditions are planned after the current items on our pipeline are implemented.
How can you visualize or represent the chemical space of a Markush structure?
Random enumeration and Markush structure reduction are useful techniques for this.
How can I identify which part of an enumerated structure comes from which R-group, etc?
The Markush enumeration plugin helps in numerous ways:
  • R-group coloring: the scaffold and all top-level R-groups are coloured differently.
  • The scaffold is oriented the same way as in the Markush structure.
  • A Markush code (enumeration ID) can describe which definition was used for the enumeration of each library members.

Try it now!

Evaluation

Need more information?

Have a look on our support forum or drop us a line.