Hi Pavel, I am very lucky to see 2 maps of the same protein, at roughly the same resolution, one by Xray and the other by EM (it's not that common). The details from the 2 maps are however different... Here is my procedure, please let me know what you think: - convert .mrc map to .mtz ; I get F, PHIF. - supperpose the maps: map1: labels 2FOFCWT, PH2FOFCWT map2: F, PHIF. I also provided the 2 PDBs as it is clearly stated on the input files definitions. It worked, with the 2 maps superposed that I can look and compare. However, for some reason, the superposition has been done on chain C and D of the crystallograhic model instead of chain AB (4 chains in the asymmetric unit, making 2 dimers). Is there a reason there, choosing the closest structure, best fit? (I see that it is possible to define a superposition) Another thing is that it creates a "supperposition in the middle", meaning neither on model1 nor model2 from the input, and not a crystallographic symmetric of model1. So I can't use the full crystallographic maps, I have a "simplified" version of it. Please let me know if I'm doing something wrong as haven't yet invested into understanding map formats. Thanks Vincent ps: Zhijie, thanks for the Chimera tip. I'll have a look into it later. On 04/10/2018 12:00, Pavel Afonine wrote:

Hi Vincent,

...

I am trying to superpose a CryoEM map onto a Xray map. I tried the "superpose maps" under the map section of phenix 1.14 but it only takes "reflections" and not the .mrc format  of the CryoEM map. should I transform the CryoEM map into structure factors using "Map to structure factors" or is there a better way to do this?

you can try.. Yes, convert both maps into Fourier space equivalents and feed into the program. Note, you need  two models that fit two maps. The way it works is that it superposes models first, then applies corresponding operators (that superpose models) to maps. I can't remember anyone asking this questions before on this list or myself directly.

Share your experience!

Good luck, Pavel

-- Vincent Chaptal, PhD MMSB -UMR5086 Drug Resistance and Membrane Proteins Laboratory 7 passage du Vercors 69007 LYON FRANCE +33 4 37 65 29 01 http://mmsb.cnrs.fr/en/

Hello I just want to follow up on this question as I have a similar one, namely I would like to align two cryo-EM maps. I have the conversion of the individual mrc files to mtz file working fine using the “map to structure factors” gui. However if I use “Superpose Maps” the resulting PDBs and maps are shifted away from the inputs. Is there anyway to have the maps in the “reference frame” of the first input model/map? I know this is easier in Chimera but I am worried that the resampling in chimera might introduce some errors that could be avoided if done in using the mtz files in Phenix (maybe this is an incorrect worry though and my question irrelevant?). Sean --------------------------------- Sean Connell, PhD CICbioGUNE Parque Tecnológico de Vizcaya, Ed. 800 48160 Derio Phone: +34 946 572 529 [email protected] www.cicbiogune.es frozenribosomes.com skype: io2-76 ---------------------------------

On 4 Oct 2018, at 16:23, vincent Chaptal wrote:

Hi Pavel,

I am very lucky to see 2 maps of the same protein, at roughly the same resolution, one by Xray and the other by EM (it's not that common). The details from the 2 maps are however different...

Here is my procedure, please let me know what you think: - convert .mrc map to .mtz ; I get F, PHIF. - supperpose the maps: map1: labels 2FOFCWT, PH2FOFCWT map2: F, PHIF. I also provided the 2 PDBs as it is clearly stated on the input files definitions.

It worked, with the 2 maps superposed that I can look and compare.

However, for some reason, the superposition has been done on chain C and D of the crystallograhic model instead of chain AB (4 chains in the asymmetric unit, making 2 dimers). Is there a reason there, choosing the closest structure, best fit? (I see that it is possible to define a superposition) Another thing is that it creates a "supperposition in the middle", meaning neither on model1 nor model2 from the input, and not a crystallographic symmetric of model1. So I can't use the full crystallographic maps, I have a "simplified" version of it.

Please let me know if I'm doing something wrong as haven't yet invested into understanding map formats. Thanks Vincent

ps: Zhijie, thanks for the Chimera tip. I'll have a look into it later.

On 04/10/2018 12:00, Pavel Afonine wrote:

...

Hi Vincent,

...

I am trying to superpose a CryoEM map onto a Xray map. I tried the "superpose maps" under the map section of phenix 1.14 but it only takes "reflections" and not the .mrc format of the CryoEM map. should I transform the CryoEM map into structure factors using "Map to structure factors" or is there a better way to do this?

you can try.. Yes, convert both maps into Fourier space equivalents and feed into the program. Note, you need two models that fit two maps. The way it works is that it superposes models first, then applies corresponding operators (that superpose models) to maps. I can't remember anyone asking this questions before on this list or myself directly.

Share your experience!

Good luck, Pavel

-- Vincent Chaptal, PhD

MMSB -UMR5086

Drug Resistance and Membrane Proteins Laboratory

7 passage du Vercors

69007 LYON

FRANCE

+33 4 37 65 29 01 tel:+33%204%2037%2065%2029%2001 http://mmsb.cnrs.fr/en/ http://mmsb.cnrs.fr/en/

_______________________________________________ phenixbb mailing list [email protected] mailto:[email protected] http://phenix-online.org/mailman/listinfo/phenixbb http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected] mailto:[email protected]

Hi Vincent,

I am very lucky to see 2 maps of the same protein, at roughly the same resolution, one by Xray and the other by EM (it's not that common). The details from the 2 maps are however different...

wow, fascinating!

Here is my procedure, please let me know what you think: - convert .mrc map to .mtz ; I get F, PHIF. - supperpose the maps: map1: labels 2FOFCWT, PH2FOFCWT map2: F, PHIF. I also provided the 2 PDBs as it is clearly stated on the input files definitions.

It worked, with the 2 maps superposed that I can look and compare.

Ok, sounds good so far. Yes, this is what I would expect!

However, for some reason, the superposition has been done on chain C and D of the crystallograhic model instead of chain AB (4 chains in the asymmetric unit, making 2 dimers). Is there a reason there, choosing the closest structure, best fit? (I see that it is possible to define a superposition)

You can specify this manually!

Another thing is that it creates a "supperposition in the middle", meaning neither on model1 nor model2 from the input, and not a crystallographic symmetric of model1. So I can't use the full crystallographic maps, I have a "simplified" version of it.

I guess this is the feature of current implementation. Pavel