From henrick@ebi.ac.uk Fri Aug 5 15:21:04 2005 Date: Wed, 3 Aug 2005 08:50:37 +0100 (BST) From: Kim Henrick To: P.J.Briggs Cc: Jawahar Swaminathan , Martyn Winn Subject: Re: CIF tokens from mtz2various Peter Sigh/groan/letting out a moan Jawahar has sent the non ccp4 tokens - many moons in the past - and anything that gets into pdbx should carry and alias to previous dictionaries but rcsb make it feel a them and us - sometimes but not making it clear the origin of nearly 1/2 the pdbx dic etc - yes of course we can use anything that is standard but _item_aliases.alias_name would help in pdbx so that existing files are handled ok the definition in pdbx dic save__refln.pdbx_F_plus _item_description.description ; The structure factor F(h,k,l) of the Friedel pair. ; _item.name '_refln.pdbx_F_plus' _item.category_id refln _item.mandatory_code no _item_type.code float save_ #--- is an incomplete version of save__refln.ccp4_SAD_F_meas_plus_au _item_description.description ; The measured value of the structure factor in arbitrary units for hkl. ; _item.name '_refln.ccp4_SAD_F_meas_plus_au' _item.category_id refln _item.mandatory_code no _item_related.related_name '_refln.ccp4_SAD_F_meas_plus_sigma_au' _item_related.function_code associated_esd _item_type.code float _item_type_conditions.code esd _item_units.code arbitrary save_ #--- as it doesnt have the esd nor does it follow the normal mmcif pattern of units i.e. it needs the au for arbitrary units - just like other tokens for structure factors and it hasnt been in the rcsb world for ages - it has been in pdbx for less than a week as our mirror update picks it up once a week and the ebi version of pdbx doesnt have this it appears to be added Fri Jul 29 11:04:31 EDT 2005 and then REMOVED as our direct copy of the official pdbx is header with Date: Mon Aug 1 18:01:17 EDT 2005 http://msdlocal.ebi.ac.uk/docs/exchange/mmcif/mmcif_pdbx.dic while this one http://mmcif.pdb.org/dictionaries/ascii/mmcif_pdbx.dic is dated Date: Fri Jul 29 11:04:31 EDT 2005 we mirror the mmcif dics via an rsync to a area that is supposed to be common by both processing sites - appears this is a different version of the pdbx dictionary than is public? bloody confusing kim On Tue, 2 Aug 2005, P.J.Briggs wrote: > > Hi Jawahar > > Further to my message from yesterday, I have spoken to someone at the > RCSB about the use of the CIF tokens from MTZ2VARIOUS. > > It appears that the PDB exchange dictionary mmcif_pdbx.dic (see e.g. > http://mmcif.rcsb.org/dictionaries/mmcif_pdbx.dic/Index/index.html) > contains equivalent tokens which the RCSB use to map the tokens output > from MTZ2VARIOUS, namely: > > _refln.ccp4_SAD_F_meas_plus_au -> _refln.pdbx_F_plus > _refln.ccp4_SAD_F_meas_plus_sigma_au -> _refln.pdbx_F_plus_sigma > _refln.ccp4_SAD_F_meas_minus_au -> _refln.pdbx_F_minus > _refln.ccp4_SAD_F_meas_minus_sigma_au -> _refln.pdbx_F_minus_sigma > _refln.ccp4_I_plus -> _refln.pdbx_I_plus > _refln.ccp4_I_plus_sigma -> _refln.pdbx_I_plus_sigma > _refln.ccp4_I_minus -> _refln.pdbx_I_minus > _refln.ccp4_I_minus_sigma -> _refln.pdbx_I_minus_sigma > _refln.ccp4_SAD_HL_A_iso -> _refln.pdbx_HL_A_iso > _refln.ccp4_SAD_HL_B_iso -> _refln.pdbx_HL_B_iso > _refln.ccp4_SAD_HL_C_iso -> _refln.pdbx_HL_C_iso > _refln.ccp4_SAD_HL_D_iso -> _refln.pdbx_HL_D_iso > > My understanding is that these are tokens which the RCSB uses for > internal purposes, however it seems sensible to update MTZ2VARIOUS to > output these tokens instead of the _refln.ccp4_* formats. > > Would this proposal be acceptable to the EBI? (It would appear to mirror > changes made previously within Refmac to use the pdbx tokens.) > > Comments welcome, thanks > > Peter > > On Mon, 1 Aug 2005, P.J.Briggs wrote: > > > > > Hi Jawahar > > > > Thanks for the information. I'm not sure what the situation is at the > > RCSB and I'm trying to find out what they do with the additional CCP4 > > tokens generally. The reference to the exchange dictionary is useful. > > > > The main issue with anomalous data is that I believe that there maybe > > some ambiguity over the use of _refln.F_meas_au and > > _refln.intensity_meas, when exporting from MTZ to CIF. In MTZ2VARIOUS > > these tokens are now used to write out the average (over Friedel mates) > > of the F and I values respectively. However when hkl and -h-k-l > > reflections are written explicitly to the mmCIF file it seems that the > > same tokens represent the F(+) value (for hkl) and F(-) value (for > > -h-k-l) (and similarly for I(+) and I(-)). > > > > Do you know do if this ambiguity is permissible? (Whether or not it is, > > it doesn't seem like a good idea to me.) > > > > Once the situation is clarified then MTZ2VARIOUS and CIF2MTZ can be > > updated to read and write the appropriate tokens. But first we need to > > know that the tokens we are reading/writing are the correct ones, for > > both the EBI and the RCSB. > > > > Thanks again and best wishes > > > > Peter > > > > On Mon, 1 Aug 2005, Jawahar Swaminathan wrote: > > > > > Dear Peter and Kim, > > > > > > These cif items occur quite often in the MTZ files we receive from our > > > depositors. > > > > > > These items are now part of the cif exchange dictionary (mmcif_ccp4.dic) > > > as we were in touch with John Westbrook more than two years ago in order > > > that we could include these columns from structure factor flags rather > > > than throwing them away. In theory, the RCSB should also be able to use > > > this information for their processing needs, since they are now part of > > > the standard data exchange mmCIF dictionaries. > > > > > > I am enclosing a copy of the mmcif_ccp4.dic file that was last updated > > > on the 25th of July. We do use cif2mtz in order that we may run SFCHECK, > > > but both cif2mtz and sfcheck throw away these items during analysis. It > > > would be great if mtz2various could parse this information and write out > > > the necessary information, when present. > > > > > > Hope that answers your question. > > > > > > best regards - Jawahar > > > > > > Jawahar Swaminathan, Ph.D. > > > PDB Depositions > > > MSD-EBI > > > > > > >>Hi Kim & Martyn > > > >> > > > >>I'm trying to unravel some issues with the tokens written out by the > > > >>MTZ2VARIOUS program. In May 2003 Martyn modified the CIF output: > > > >> > > > >> > > > >> > > > >>The explicit anomalous columns/tokens are: > > > >> > > > >> _refln.ccp4_SAD_F_meas_plus_au F(+) > > > >> _refln.ccp4_SAD_F_meas_plus_sigma_au SIGF(+) > > > >> _refln.ccp4_SAD_F_meas_minus_au F(-) > > > >> _refln.ccp4_SAD_F_meas_minus_sigma_au SIGF(-) > > > >> _refln.ccp4_SAD_phase_anom DP > > > >> _refln.ccp4_SAD_phase_anom_sigma SIGDP > > > >> _refln.ccp4_I_plus I(+) > > > >> _refln.ccp4_I_plus_sigma SIGI(+) > > > >> _refln.ccp4_I_minus I(-) > > > >> _refln.ccp4_I_minus_sigma SIGI(-) > > > >> > > > >>There are some problems: > > > >> > > > >>1. CIF2MTZ doesn't recognise these so it cannot back-transform the CIF > > > >> output from MTZ2VARIOUS to MTZ (this is not really of interest here) > > > >>2. The RCSB doesn't recognise these tokens and so MTZ2VARIOUS > > > >> cannot be used to generate structure factors for deposition if > > > >> anomalous data is present. > > > >> > > > >>So what I want to know is: does the EBI recognise/accept these tokens in > > > >>deposited CIF files? or are they converted to something else, or what? > > > >> > > > >>Thanks for your help, best wishes > > > >> > > > >>Pete > > > >> > > > >>-- > > > >>_____________________________________________________ > > > >>Peter J Briggs, pjx@ccp4.ac.uk Tel: +44 1925 603826 > > > >>CCP4, ccp4@ccp4.ac.uk Fax: +44 1925 603825 > > > >> http://www.ccp4.ac.uk/ > > > >>Daresbury Laboratory, Daresbury, Warrington WA4 4AD > > > >> > > > >> > > > >> > > > > > > > >Kim HENRICK henrick@ebi.ac.uk ::telephone: +44 (0) 1223 494629 > > > > > > > > > > > > > > > > > > > > > > -- > > _____________________________________________________ > > Peter J Briggs, pjx@ccp4.ac.uk Tel: +44 1925 603826 > > CCP4, ccp4@ccp4.ac.uk Fax: +44 1925 603825 > > http://www.ccp4.ac.uk/ > > Daresbury Laboratory, Daresbury, Warrington WA4 4AD > > > > > > -- > _____________________________________________________ > Peter J Briggs, pjx@ccp4.ac.uk Tel: +44 1925 603826 > CCP4, ccp4@ccp4.ac.uk Fax: +44 1925 603825 > http://www.ccp4.ac.uk/ > Daresbury Laboratory, Daresbury, Warrington WA4 4AD > Kim HENRICK henrick@ebi.ac.uk ::telephone: +44 (0) 1223 494629