<Minor late night typo edit and update. This story took place more than 13 years ago, before the current big push towards open data really took off>
A few years ago we published a paper on the NMR solution structures of some small peptides. I won’t post the link to the the actual paper because it’s not really relevant to today’s story.
NMR solution structures of short peptides are funny sort of things. Unlike small molecule crystal structures, where on publication you’re expected to lodge the structures into a database like the CCDC – Cambridge Crystallographic Database, or protein structures where you’re expected to lodge the structures in the PDB (and most journals require you to do so along with publication), small peptide NMR structures are not allowed in the PDB. Currently the minimum length accepted is 24 residues. The reason for this is that small peptides often adopt a range of structures in solution, and hence the PDB moderators reason that such structures are often too imprecise to go in a scholarly database. There are points to be made on both sides of this debate and I acknowledge that this is a valid concern. Frustrating, as we’re about to see, but I gladly accept their decision. One alternative is to publish the NMR structures into a third database, the BMRB – Biological Magnetic Resonance Bank. This repository allows the upload of NMR constraint data as well, so people can in theory at least reproduce published results, or at least run the same data through their preferred structure calculation program.
The paper we published in did not require us to include the structures in the Supporting Material and so we didn’t. The molecules were also covered by a patent issued to the University and the industrial sponsor of the research, complicating matters slightly. Prior to publication we had been contacted a few times by competing groups in the field asking us if we had NMR structures for these peptides and if they could get access to them. Due to the emerging patent situation we declined. After publication those requests dried up. However one of our competitors did publish a paper a couple of years later, publicly berating us for not releasing the data into the public domain. They explicitly commented that by not releasing the data we were actively hindering the field of research.
The crux of their “new” paper was basically “we had to repeat the work and here is the same structure.” The ONLY point of novelty of their “work” was that they claim to have identified under the same conditions as us, the presence of a second small population conformer of the peptide.
They were wrong. If they had read our other papers on the subject they would have known that the synthesis of this molecule occasionally generates a small amount of epimerisation at the C-terminus of the peptide. What they were seeing as a “new” conformer was in fact merely a diasteomeric impurity. The compound itself is now available from commercial vendors. I know that early batches were plagued with this impurity at abundances up to 50%. We’ve always made our own.
Anyway they went off in a huff and “published” their “new” “public” structure. And guess what, they NEVER published the coordinates or even the constraints files. Not in BRMB. Not in PDB. Not anywhere. Nice.
<update: I double checked again this evening and their data is still not in BRMB where they intended to put it>