Re: [COOT] Map rotation and translation
Thank you Paul for the explanation/clarification. Indeed, it would be enough for me at the moment if I could align maps according to the fit I get in ChimeraX. The "transform_map() function" sounds exactely like what I was looking for! Could you perhaps elaborate if I filled in the parameters correctly (see below)? ChimeraX output: Fit map job122_class1 in map job145_class1+3 using 378391 points correlation = 0.9626, correlation about mean = 0.8182, overlap = 1.076e+04 steps = 40, shift = 0.009, angle = 0.00377 degrees Position of job122_class1 (#1) relative to job145_class1+3 (#4) coordinates: Matrix rotation and translation 0.3251 -0.00473394 -0.01060925 2.07918736 0.00471099 0.8651 -0.00218664 -0.57281865 0.01061946 0.00213651 0.4133 -1.65452589 Axis 0.18292243 -0.89823687 0.39963723 Axis point 165.64873479 0. 190.89655799 Rotation angle (degrees) 0.67707290 Shift along axis 0.23364667 Coot command: Based on |(transform-map /imol/ /rotation-matrix/ /trans/ /point radius/)| from the user manual: (transform-map 1 (list 0.3251 -0.00473394 -0.01060925 0.00471099 0.8651 -0.00218664 0.01061946 0.00213651 0.4133) (list 2.07918736 -0.57281865 -1.65452589) (list 181.1739959716797 205.94700622558594 198.18699645996094) 10) The value for "point" I took from running |(rotation-centre)| in Coot, is that correct? The resulting map fragment seems to be aligned quite well to the resampled map.| | I set the radius to 10 for testing purposes, is there a way to use the full map by default? Best, Helge To unsubscribe from the COOT list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=COOT&A=1 This message was issued to members of www.jiscmail.ac.uk/COOT, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
Re: [COOT] Map rotation and translation
Paul, both your emails are partly correct. Rotation+translation, called skew transformation (words 25-37) is supported by CCP4 maplib and PyMOL. It's mentioned in the MRC/CCP4 2000 specification on CCP-EM page: https://www.ccpem.ac.uk/mrc_format/mrc2000.php although it's not mentioned in the 2014 version of this spec. I don't know if any program apart from PyMOL is using it. When Tom wrote about the shift of origin, he must have meant N[XYZ]START fields. Marcin On Fri, Jan 28, 2022 at 6:00 AM Paul Emsley wrote: > > On 27/01/2022 01:13, Paul Emsley wrote: > > > > Alas, I have not done what I think you want (which is to say, to use > > the extented CCP4/MRC map header information to represent an arbitrary > > rotation/translation). > > > > After having read Tom's reply about these > rotation/translation/interpolation issues on the ChimeraX mailing list > I'd like to correct a misunderstanding I had about what Coot wasn't > doing. There is an unused (by Coot) slot in the CCP4/MRC map header, I > had thought that the slot allows rotation and translation but it does > not - it allows a shift of origin (translation) only. Everything Tom > said was on the money. > > Paul. > > > > To unsubscribe from the COOT list, click the following link: > https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=COOT&A=1 > > This message was issued to members of www.jiscmail.ac.uk/COOT, a mailing list > hosted by www.jiscmail.ac.uk, terms & conditions are available at > https://www.jiscmail.ac.uk/policyandsecurity/ To unsubscribe from the COOT list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=COOT&A=1 This message was issued to members of www.jiscmail.ac.uk/COOT, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
Re: [COOT] Map rotation and translation
On 27/01/2022 01:13, Paul Emsley wrote: Alas, I have not done what I think you want (which is to say, to use the extented CCP4/MRC map header information to represent an arbitrary rotation/translation). After having read Tom's reply about these rotation/translation/interpolation issues on the ChimeraX mailing list I'd like to correct a misunderstanding I had about what Coot wasn't doing. There is an unused (by Coot) slot in the CCP4/MRC map header, I had thought that the slot allows rotation and translation but it does not - it allows a shift of origin (translation) only. Everything Tom said was on the money. Paul. To unsubscribe from the COOT list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=COOT&A=1 This message was issued to members of www.jiscmail.ac.uk/COOT, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
Re: [COOT] Map rotation and translation
On 27/01/2022 01:13, Paul Emsley wrote: I would like to draw your attention to the following paper and the EMDA project by my colleagues, it is a cryo-EM map and model manipulation toolbox, maybe that has what you need: https://www.biorxiv.org/content/10.1101/2021.07.26.453750v2 This has been published now: https://www.sciencedirect.com/science/article/pii/S1047847721001313 To unsubscribe from the COOT list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=COOT&A=1 This message was issued to members of www.jiscmail.ac.uk/COOT, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
Re: [COOT] Map rotation and translation
On 26/01/2022 22:18, Helge Paternoga wrote: Dear all, Continuing from this topic over at the ChimeraX user list: http://plato.cgl.ucsf.edu/pipermail/chimerax-users/2022-January/003177.html I would like to ask if it is possible currently (or to introduce) the ability in Coot to rotate and translate a cryo-EM map according to the output of the ChimeraX "fitmap" command. This would be very useful in comparing the map quality of volumes e. g. from the same classification. Resampling the volumes in ChimeraX changes the map and is therefore not a good option when one wants to compare subtle differences in the densities. I believe the information provided by "fitmap" adheres to the ChimeraX Positions Format described here: https://rbvi.ucsf.edu/chimerax/docs/user/formats/positions.html The output from "fitmap" to the log appears a bit different though, so I am not sure which one would be useable in the end: Example: tool show 'Fit in Map'Fit map run_class001.mrc in map run_class001.mrc using 159932 points correlation = 0.9589, correlation about mean = 0.5393, overlap = 82.65 steps = 48, shift = 2.47, angle = 0.571 degrees Position of run_class001.mrc (#1) relative to run_class001.mrc (#2) coordinates: Matrix rotation and translation 0.5193 -0.00918946 0.00341831 2.19650599 0.00918323 0.5615 0.00183397 -4.59169731 -0.00343502 -0.00180249 0.9248 1.52902230 Axis -0.18233729 0.34363524 0.92123175 Axis point 479.76292846 225.20927991 0. Rotation angle (degrees) 0.57135182 Shift along axis -0.56979010 Alas, I have not done what I think you want (which is to say, to use the extented CCP4/MRC map header information to represent an arbitrary rotation/translation). In my tests, transforming the map changes the values by 0.1% or less and that is often acceptable it seems. You could validate your results by making the transformation the other way around. You could use refmac to calculate the fourier coefficients and write an mtz file. On reading that into Coot you can control the Shannon super-sampling rate and that will reduce the interpolation errors. To apply the transformation above to a map in Coot, one would use the function transform_map() function: # transform_map(imol, mat, trans, about_pt, radius, space_group, cell) # # where space_group is a "P1" and cell is a list of 6 parameters, where the cell angles are in degrees (you can use the input cell) You can access the cell parameters using the cell() function. The mat is a 3x3 matrix list and the trans in this case is [2.1965059, -4.5916973, 1.5290223] I would like to draw your attention to the following paper and the EMDA project by my colleagues, it is a cryo-EM map and model manipulation toolbox, maybe that has what you need: https://www.biorxiv.org/content/10.1101/2021.07.26.453750v2 Paul To unsubscribe from the COOT list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=COOT&A=1 This message was issued to members of www.jiscmail.ac.uk/COOT, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
