argg .. that is annoying - Pedro Beltrao

Have you ever tried to get access to these private datasets on a collaboration basis ? - Pedro Beltrao

GSK has a matrix of a couple thousand compounds versus 800 assays, heard that in a talk ~1.5 years ago. - Josef Scheiber

To add and shamelessly cite myself (http://pubs.acs.org/doi...): About two years ago we were dealing with roughly 2.4 million unique datapoints at Novartis (which includes several commercial DBs). Now they are dealing with 20 million datapoints (http://www.bio-itworldexpo.com/Bio-It_... ==> abstract from Flo Nigsch) - Josef Scheiber

At a startup I was at, we were trying to find more targets for pharma. Turned out they already had more than they could handle, and this was 10 years ago. - Deepak Singh

+1 Josef - When looking @PubChem BioAssay http://www.ncbi.nlm.nih.gov/pcassay I am seriously wondering about data consistency and assay screening conditions? What are we looking at here? apples or oranges? - joergkurtwegner

@joerg, apples and oranges :( - Rajarshi Guha

I just heard from some people that using PubChem BioAssay might be a reason to get a publication killed by reviewers, did anyone else experience this already? - joergkurtwegner

@Joerg, Rajarshi, probably also some other fruit ;-) But there's a nice initiative under way to clean this up: http://bioassayontology.org/ - Josef Scheiber

@Sepp, yes, this looks interesting - while it will certainly help standardization, i'm not sure that it will address the actual data quality itself - Rajarshi Guha

Hehe, I'm a hopeless optimist ;-) I'm on the standpoint that it will help identifying errors sooner and therefore enhance the quality significantly on the long term, until then we'll have to live with the stuff in there ... - Josef Scheiber