On Mon, 07 Mar 2016, Peter Rice wrote: > We had this discussion some years back about SwissProt protein > sequence entries included as test data in EMBOSS. We also have PDB > files in the EMBOSS test data. > > The conclusion was that scientific data (SwissProt, PDB, etc.) are > scientific facts and it is not reasonable to require permission to > change them.
This isn't true; there are loads of reasons to change sequences and structural models of proteins. Protein sequences are just based on references which have inaccuracies and do not represent ancestral sequences or the true variation present in real populations; in my lab we modify UniProt sequences and redistribute those modifications in publications all of the time. PDB files are also just the current model of a particular crystallization of a particular set of molecules. It's pretty standard practice to modify them using simulations to account for different binding partners and assemblies. [I've published the PDB files containing these modifications, for example.] All of that said, because PDB and UniProt files are not works of authorship, they likely do not qualify for copyright protection in the US, so the licensing terms can largely be ignored. However, that may not be true of other jurisdictions. -- Don Armstrong http://www.donarmstrong.com It is easier to build strong children than to repair broken men. -- Frederick Douglass