Pfizer Selects ChemAxon's Markush Platform
Pfizer Globally Licenses ChemAxon's Platform For Markush Structure Searching And Enumeration
ChemAxon - a leading provider of cheminformatics software for the life sciences industry has entered into agreement with Pfizer to provide its Markush searching and structure enumeration functionality for global development and deployment within Pfizer's global R&D facilities. "Markush structure search and enumeration is becoming of critical importance for structure IP and patent development." said Alex Drijver, CEO of ChemAxon. "It is an area where ChemAxon has developed a high area of expertise and where we are moving very much into the forefront."
ChemAxon is a leader in providing cheminformatics software development platforms and applications for the biotechnology, pharmaceutical and agrochemical industries. With core capabilities for structure visualization, search and management, property prediction, virtual synthesis, screening and drug design, ChemAxon focuses upon active interaction with users and software portability to create powerful, cost effective cross platform solutions and programming interfaces to power modern cheminformatics and chemical communication.
This is remarkable news. It means that at least one of major companies moved forward for in-house Markush patent processing and searching.
To understand importance of this event we need to collect several known pieces of information together.
1. ChemAxon - a leading software company developing cheminformatics toolkits and applications. It has been involved in development software for handling Markush structures for a number of years. The last presentation of Szabolcs Csepregi at ICIC-2009 Markush structures – From molecules towards patents summarizes state of work of ChemAxon on handling Markush Structures. It was announced during the talk and made news on Twitter, that ChemAxon's MarvinSketch 5.3 would be able to import Markush structures from DWPI in .VMN (Markush DARC format) files (see slide 10 of the talk). Marvin 184.108.40.206 was released (January 21th, 2010), and at the same day ChemAxon announced its global license to Pfizer.
2. Anton Fliri from Pfizer (US) reported at ICIC-2008 (Nice, France, Oct. 2008) on evaluation of Markush structure enumeration technology in collaborations with INPI and Decript. In this presentation, Dr Fliri described technique of ranking patent search results (found, for example, in SciFinder using "Like" query) by similarity of compounds enumerated from topological descriptors of Markush structures from patents in the search result list. Top ranked patent are found highly relevant. It was shown (see Slide 12) that ranking patents based only on specific compounds did not provide satisfactory results (Note: MMS database has limited number compounds from patents; use of all prophetic compounds from patents could potentially gives comparable results) Details of Use enumeration technology for processing Markush structures disclosed in US20090132464. Note that technology starts from initial database search which according to Claim 3, could be done in MMS or MARPAT database.
3. It is interesting to analyze assignment information for US20090132464. The application has four inventors. The first inventor Dr Fliri, reassigned its right to Pfizer, which in turn reassigned them to Decript. Three other inventors reassigned their rights to directly to Decript. I suggest that Decript was founded handle further technology transfer inside The Pistoia Alliance, which was "established to streamline non-competitive elements of the pharmaceutical drug discovery workflow". Structure Search and Markush representation for patent data are among current projects. Decript technology could be further transferred to other The Pistoia Alliance members, which include, beyond Pfizer, also AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, GlaxoSmithKline, Hoffmann-La Roche, Lundbeck, Merck and Novartis. Decript may be presented in the Alliance through ChemITment, one of the founding members, which related to Decript through a ex-Pfizer inventor of US20090132464. (Chemaxon and Thomson Reuters are member of the Alliance as well).
Licensing ChemAxon Markush platform, using Markush topological data from MMS database and having Decript technology, Pfizer get a powerful tool which enable Pfizer scientists address new paradigm: "The Burden Of Proof That Inventions Are Novel And Patentable is Shifting To Inventors" (see Anton Fliri talk, slide 4). ('Fail Fast, Fail Cheap' mantra?). Other pharmaceutical companies, at least from The Pistoia Alliance, could benefit from this development too.
Does it mean that scientists would be now able to get good enough results on novelty evaluation without professional patent searcher? I think so. A scientist will 1be able to identify relevant patents, if any, much faster than now, using such or similar technology.
What it means if no relevant patent have been found? We all know the answer - nothing: every search system has its inherit limitations.
I believe that for critical projects expert searchers would be always involved, and they would benefit if routine cases with simple answers would be screen out by scientists themselves.
Regards, Aleksandr Belinskiy Patent Information Consultant