The unacceptable state of Chemoinformatics

Egon Willighagen has cricitized a numner of aspects of chemoinformatics: (I don’t blame Individuals in Commercial Chemoinformatics). This post was sparked off by an announcement that company A had agreed with company B that A would use B’s software. This touched a raw nerve in Egon who asked – why is this such a big deal? I understand Egon’s position – I have had 25+ years of these marketing announcements – X and Y announce some joint activity and I’m now inured to that. But he’s right – this is not news – it’s marketing.
Ultimately I blame the pharma companies – they have abrogated any sense of scientific or other rigour in chemoinformatics yet they effectively dominate the subject. Work in chemoinformatics is slanted towards what will appeal to pharma (and the appropriate publication mechanisms) rather than advancing our understanding of the science. The average publication relies on unobtainable datasets, closed bespoke software, undisclosed methodology and therefore unrepeatable science. It’s not surprising that we see the subject contracting – witness the closure of Cologne and the failure to fund the NIH Chemoinformatics program.
The pharma industry actively works against Open processes. Their exploratory science is in crisis. They would benefit from ODOSOS – Open Data (e.g. in basic chemistry, chemoinformatics, biological activity, ADME, pharmacology, etc.) Open Source and Open Standards. Yet they hide from sight, while actually using considerable amounts of Open Source software. and data. While this is not illegal, it’s certainly questionable whether an industry should take so much without returning anything – if nothing else on simple utilitarian grounds.
I’m feeling personally abused having been let down by a pharma company who approached me saying they were interested in Open Source software, wanted me to visit and then let me down. It wasted my time and left me out of pocket. It’s not the first time. The Blue Obelisk has now good products – Open Babel, CDK, Jmol, CML, etc. We know they are used in companies. It would be nice to feel this was valued. They could afford to feed something back.