Re: [ccp4bb] How to make covalent bond between sugar monomers?
Hi Leo, you should not need an additional cif file for coot but you might need to generate one for reciprocal space refinement. If you use REFMAC it might 'fail' in the first attempt with the message 'new ligand encountered' or something similar. It will, however, generate a single cif file for the sugar if it recognizes the monosaccharides and the glycosidic linkage you have created in coot, using the descriptions from the REFMAC library. You can then look at this cif file and provide it for future refinement runs in REFMAC. It is a good idea to keep checking if the glycosidic bond is identified correctly and restrained accordingly in every REFMAC run after you introduced the ligand/glycosylation. I find it a lot harder to achieve correct sugar refinement with phenix, although elbow should techically be able to create a cif file for sugars just like REFMAC. The prodrug server is a bad idea, as pointed out, because the library restraints for sugars are not used if the sugar rings are not identified with the correct three letter abbrevations in your pdb file. If you have a larger oligosaccharide and good density it might be easiest to start by getting a similar sugar model from a deposited pdb file. Place this molecule in your sugar density using rigid body refinement, then start to manipulate single rings from the template so that they fit your density best. Often this is a matter of rotating the individual rings around their glycosidic bonds. When you try this procedure be aware that there are lots of distorted sugar rings in the pdb, so make sure you start from a good template. Good luck, Baerbel Zitat von Lie Ma leoma5...@gmail.com: Thank you, Paul. And do I need to provide additional cif file which describe the bond? Leo On Sep 4, 2014, at 2:23 PM, Paul Emsley pems...@mrc-lmb.cam.ac.uk wrote: On 04/09/14 19:06, Lie Ma wrote: Hello all, I am trying to make a model of an oligosaccharide which contains several sugar monomers. I obtained the monomers from coot library and put them into the electron density. But I don't know how to make covalent bond between these monomers. If you are using Coot, you can use Extensions - Modelling - Make Link. If you sugar monomers are reasonably placed Coot's Sphere Refine will determine the correct link types on the fly (it doesn't believe what's in the PDB file). I tried to use PRODRG to generate a model and it gave me one big ligand consisting of several monomers. Yes, that's not the way to do it... However, what I want is to have a chain contains several monomers, and each monomers acts like a residue. Indeed. Paul. -- Bärbel Blaum, Ph.D. Interfakultäres Institut für Biochemie (IFIB) Hoppe-Seyler-Strasse 4 D-72076 Tübingen Germany +49 70 71 29 73 375
[ccp4bb] AW: [ccp4bb] How to make covalent bond between sugar monomers?
Dear Leo, Just in case, you have to delete the oxygen from the carbon making the covalent link in the sugar you are adding to the chain (in the direction starting from the Asn or Ser). This is usually the O1. Then you make a link between this carbon and the connecting O or N atom. I usually do a real space refinement of the single sugar unit and I continue rebuilding until the atoms to be connected are within reasonable distance to make the link after this real space refinement. You may have to rebuild the Asn or Ser as well. In my experience, if the link is correctly specified and the sugar is correctly modeled, it will refine correctly without having to add or generate an additional cif file. Good luck, Herman -Ursprüngliche Nachricht- Von: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] Im Auftrag von Bärbel Blaum Gesendet: Freitag, 5. September 2014 08:49 An: CCP4BB@JISCMAIL.AC.UK Betreff: Re: [ccp4bb] How to make covalent bond between sugar monomers? Hi Leo, you should not need an additional cif file for coot but you might need to generate one for reciprocal space refinement. If you use REFMAC it might 'fail' in the first attempt with the message 'new ligand encountered' or something similar. It will, however, generate a single cif file for the sugar if it recognizes the monosaccharides and the glycosidic linkage you have created in coot, using the descriptions from the REFMAC library. You can then look at this cif file and provide it for future refinement runs in REFMAC. It is a good idea to keep checking if the glycosidic bond is identified correctly and restrained accordingly in every REFMAC run after you introduced the ligand/glycosylation. I find it a lot harder to achieve correct sugar refinement with phenix, although elbow should techically be able to create a cif file for sugars just like REFMAC. The prodrug server is a bad idea, as pointed out, because the library restraints for sugars are not used if the sugar rings are not identified with the correct three letter abbrevations in your pdb file. If you have a larger oligosaccharide and good density it might be easiest to start by getting a similar sugar model from a deposited pdb file. Place this molecule in your sugar density using rigid body refinement, then start to manipulate single rings from the template so that they fit your density best. Often this is a matter of rotating the individual rings around their glycosidic bonds. When you try this procedure be aware that there are lots of distorted sugar rings in the pdb, so make sure you start from a good template. Good luck, Baerbel Zitat von Lie Ma leoma5...@gmail.com: Thank you, Paul. And do I need to provide additional cif file which describe the bond? Leo On Sep 4, 2014, at 2:23 PM, Paul Emsley pems...@mrc-lmb.cam.ac.uk wrote: On 04/09/14 19:06, Lie Ma wrote: Hello all, I am trying to make a model of an oligosaccharide which contains several sugar monomers. I obtained the monomers from coot library and put them into the electron density. But I don't know how to make covalent bond between these monomers. If you are using Coot, you can use Extensions - Modelling - Make Link. If you sugar monomers are reasonably placed Coot's Sphere Refine will determine the correct link types on the fly (it doesn't believe what's in the PDB file). I tried to use PRODRG to generate a model and it gave me one big ligand consisting of several monomers. Yes, that's not the way to do it... However, what I want is to have a chain contains several monomers, and each monomers acts like a residue. Indeed. Paul. -- Bärbel Blaum, Ph.D. Interfakultäres Institut für Biochemie (IFIB) Hoppe-Seyler-Strasse 4 D-72076 Tübingen Germany +49 70 71 29 73 375
Re: [ccp4bb] Extract Euler angles from fractional coordinate matrix
Rotation matrices are rarely specified in a fractional coordinate system? The criteria for checking such a matrix is Is the determinant 1? and this only holds for orthogonal matrices. I guess the way I would do this though. You presumably have two sets of fractional coordinates, before and after rotation? There is a ccp4 program - coordconv which will read the fractional coordinates and generate pdb format with the convention ncode = 1 (You may need to fudge the fractional format I suppose..) You can then use superpose to match the two sets of coordinates and the output will tell you the Eulerian angles used for the rotation! Lots of ways to kill cats! Eleanor On 4 September 2014 21:21, Phil Jeffrey pjeff...@princeton.edu wrote: The orthogonal/fractional matrix is outlined here: http://www.iucr.org/__data/assets/pdf_file/0009/7011/19_ 06_cowtan_coordinate_frames.pdf Sorry to say I apparently ditched my old Fortran o2f and f2o programs to do that. Bear in mind, however, that orthogonal has no fixed orientation with respect to fractional - for most space groups ncode 1 is often used but for primitive monoclinic ncode 3 is sometimes used, and I think the matrix shown in Kevin Cowtan's document above corresponds to ncode 1. Phil Jeffrey Princeton On 9/4/14 3:55 PM, Chen Zhao wrote: I am sorry, just to clarify, the fractional coordinate matrix I referred to is a rotational matrix in the fractional coordinate system. On Thu, Sep 4, 2014 at 3:52 PM, Chen Zhao c.z...@yale.edu mailto:c.z...@yale.edu wrote: Hi all, I am just curious whether there are some tools extracting the Euler angles from a fractional coordinate matrix. I have no luck searching it online. Alternatively, I found the analytical solution for the Euler angles from an orthogonal coordinate matrix. So in the worst case, my problem reduces to calculating the transformation matrix between the fractional and orthogonal coordinate system. I feel a little bit at a loss because it is 6 years since I last studied linear algebra. How can I calculate this for a specific unit cell? Thanks a lot in advance! Sincerely, Chen
Re: [ccp4bb] Reliable criteria to tell Anomalous or not?
Useable anomalous signal depends on what you want to use it for... If it is to generate experimental phases, then the SHELXC/D/E criteria of Dano/SigDano are pretty good. If you already know the structure and want to check if there is Br there then a much weaker signal can answer the question. You need to do a anomalous difference map and search for peaks.. Steps are: Run refmac to generate output file with h k l F SIGF ... PHIC .. FOM.. Run cad to merge refmac output with the DANO SIGDANO columns generated by Ctruncate after data processing runn fft with the anomalous map option and a peak search. Inputs DANO SIDANO PHIC FOM ( fft automatically advances PHIC by 90 degrees ) (Yes - it SHOULD be automated!! ) Read the peak search into coot Check whether the sites show up likely anomalous scatterers - eg Br, S, Ca etc.. This often works Eleanor On 4 September 2014 22:47, Boaz Shaanan bshaa...@bgu.ac.il wrote: Also, if you process your data using XDS, it'll give you good indication for the strength of the anomalous signal. I find the XDS and Aimless indications to agree quite well. Boaz *Boaz Shaanan, Ph.D. Dept. of Life Sciences Ben-Gurion University of the Negev Beer-Sheva 84105 Israel E-mail: bshaa...@bgu.ac.il bshaa...@bgu.ac.il Phone: 972-8-647-2220 Skype: boaz.shaanan Fax: 972-8-647-2992 or 972-8-646-1710* -- *From:* CCP4 bulletin board [CCP4BB@JISCMAIL.AC.UK] on behalf of CPMAS Chen [cpmas...@gmail.com] *Sent:* Thursday, September 04, 2014 11:05 PM *To:* CCP4BB@JISCMAIL.AC.UK *Subject:* [ccp4bb] Reliable criteria to tell Anomalous or not? Dear All CCP4 users, Recently, I have protein crystals cocrytallized with Br-containing ligand. sometimes, I feel confusion about whether the data have reliable anomalous information or not. Do you guys have some recommendation of the criteria? phenix reported anomalous measurability, CCP4/aimless has RCRanom. Sometimes, they are not consistent. Thanks! Charles -- *** Charles Chen Research Associate University of Pittsburgh School of Medicine Department of Anesthesiology **
Re: [ccp4bb] Extract Euler angles from fractional coordinate matrix
This is just a petty function in CCP4's MMDB library -- Eugene On 4 Sep 2014, at 20:52, Chen Zhao wrote: Hi all, I am just curious whether there are some tools extracting the Euler angles from a fractional coordinate matrix. I have no luck searching it online. Alternatively, I found the analytical solution for the Euler angles from an orthogonal coordinate matrix. So in the worst case, my problem reduces to calculating the transformation matrix between the fractional and orthogonal coordinate system. I feel a little bit at a loss because it is 6 years since I last studied linear algebra. How can I calculate this for a specific unit cell? Thanks a lot in advance! Sincerely, Chen -- Scanned by iCritical.
Re: [ccp4bb] Extract Euler angles from fractional coordinate matrix
Thank you Eleanor for your reply. I am actually considering how to describe a pseudo-NCS with an arbitrary rotational and translational relationship. I don't have to do this but I am just curious. It is more straightforward if I say how the two molecules are related by a rotation around unit cell axis than around orthogonal coordinate axis, which does not have an absolute physical meaning. The command line output after coot superpose prints out the rotational and translational matrices for both the orthogonal and fractional coordinate system. For using coordconv, my concern is that if I deal with a low-symmetry unit cell, which is not orthogonal by itself, the Euler angles for the fractional coordinate system and the orthogonal coordinate system should be different. If I just feed some numbers into coordconv, will it consider them as orthogonal coordinates? Thank you, Chen On Fri, Sep 5, 2014 at 6:24 AM, Eleanor Dodson eleanor.dod...@york.ac.uk wrote: Rotation matrices are rarely specified in a fractional coordinate system? The criteria for checking such a matrix is Is the determinant 1? and this only holds for orthogonal matrices. I guess the way I would do this though. You presumably have two sets of fractional coordinates, before and after rotation? There is a ccp4 program - coordconv which will read the fractional coordinates and generate pdb format with the convention ncode = 1 (You may need to fudge the fractional format I suppose..) You can then use superpose to match the two sets of coordinates and the output will tell you the Eulerian angles used for the rotation! Lots of ways to kill cats! Eleanor On 4 September 2014 21:21, Phil Jeffrey pjeff...@princeton.edu wrote: The orthogonal/fractional matrix is outlined here: http://www.iucr.org/__data/assets/pdf_file/0009/7011/19_ 06_cowtan_coordinate_frames.pdf Sorry to say I apparently ditched my old Fortran o2f and f2o programs to do that. Bear in mind, however, that orthogonal has no fixed orientation with respect to fractional - for most space groups ncode 1 is often used but for primitive monoclinic ncode 3 is sometimes used, and I think the matrix shown in Kevin Cowtan's document above corresponds to ncode 1. Phil Jeffrey Princeton On 9/4/14 3:55 PM, Chen Zhao wrote: I am sorry, just to clarify, the fractional coordinate matrix I referred to is a rotational matrix in the fractional coordinate system. On Thu, Sep 4, 2014 at 3:52 PM, Chen Zhao c.z...@yale.edu mailto:c.z...@yale.edu wrote: Hi all, I am just curious whether there are some tools extracting the Euler angles from a fractional coordinate matrix. I have no luck searching it online. Alternatively, I found the analytical solution for the Euler angles from an orthogonal coordinate matrix. So in the worst case, my problem reduces to calculating the transformation matrix between the fractional and orthogonal coordinate system. I feel a little bit at a loss because it is 6 years since I last studied linear algebra. How can I calculate this for a specific unit cell? Thanks a lot in advance! Sincerely, Chen
[ccp4bb] refine structure with mtz file in P212121 or P222?
Hi all, I am working with a protein-ligand complex structure. The data was indexed with XDS and mtz file is generated using Phenix---Reflection tools---Reflection file editor. The space group used was P222. Then, I used the Phenix---Molecular replacement to find a solution (try all possible in same pointgroup), and the MR solution was in space group P212121. And then, I reindexed the data in XDS using space group P212121 and then generate the mtz file in space group P212121. So, I have two mtz file, in space group P212121 and P222. When I refine the MR solution with the different mtz files, the map generated is not same for the ligand. The statisitcs is better with P222 mtz file than P212121 mtz file. I am wondering why the density for the ligand and the statistics is different using mtz file in different space groups and which mtz file should I use to get the final structure model. Thank you so much! Best, Wei
Re: [ccp4bb] Extract Euler angles from fractional coordinate matrix
I don't think Eulerian angles are defined for a non-orthogonal axis system? ? They are defined relative to perpendicular axes X Y Z e.g. Rotate coordinates by gamma about Z, beta about Y', alpha about Z. Eleanor On 5 September 2014 16:27, Chen Zhao c.z...@yale.edu wrote: Thank you Eleanor for your reply. I am actually considering how to describe a pseudo-NCS with an arbitrary rotational and translational relationship. I don't have to do this but I am just curious. It is more straightforward if I say how the two molecules are related by a rotation around unit cell axis than around orthogonal coordinate axis, which does not have an absolute physical meaning. The command line output after coot superpose prints out the rotational and translational matrices for both the orthogonal and fractional coordinate system. For using coordconv, my concern is that if I deal with a low-symmetry unit cell, which is not orthogonal by itself, the Euler angles for the fractional coordinate system and the orthogonal coordinate system should be different. If I just feed some numbers into coordconv, will it consider them as orthogonal coordinates? Thank you, Chen On Fri, Sep 5, 2014 at 6:24 AM, Eleanor Dodson eleanor.dod...@york.ac.uk wrote: Rotation matrices are rarely specified in a fractional coordinate system? The criteria for checking such a matrix is Is the determinant 1? and this only holds for orthogonal matrices. I guess the way I would do this though. You presumably have two sets of fractional coordinates, before and after rotation? There is a ccp4 program - coordconv which will read the fractional coordinates and generate pdb format with the convention ncode = 1 (You may need to fudge the fractional format I suppose..) You can then use superpose to match the two sets of coordinates and the output will tell you the Eulerian angles used for the rotation! Lots of ways to kill cats! Eleanor On 4 September 2014 21:21, Phil Jeffrey pjeff...@princeton.edu wrote: The orthogonal/fractional matrix is outlined here: http://www.iucr.org/__data/assets/pdf_file/0009/7011/19_ 06_cowtan_coordinate_frames.pdf Sorry to say I apparently ditched my old Fortran o2f and f2o programs to do that. Bear in mind, however, that orthogonal has no fixed orientation with respect to fractional - for most space groups ncode 1 is often used but for primitive monoclinic ncode 3 is sometimes used, and I think the matrix shown in Kevin Cowtan's document above corresponds to ncode 1. Phil Jeffrey Princeton On 9/4/14 3:55 PM, Chen Zhao wrote: I am sorry, just to clarify, the fractional coordinate matrix I referred to is a rotational matrix in the fractional coordinate system. On Thu, Sep 4, 2014 at 3:52 PM, Chen Zhao c.z...@yale.edu mailto:c.z...@yale.edu wrote: Hi all, I am just curious whether there are some tools extracting the Euler angles from a fractional coordinate matrix. I have no luck searching it online. Alternatively, I found the analytical solution for the Euler angles from an orthogonal coordinate matrix. So in the worst case, my problem reduces to calculating the transformation matrix between the fractional and orthogonal coordinate system. I feel a little bit at a loss because it is 6 years since I last studied linear algebra. How can I calculate this for a specific unit cell? Thanks a lot in advance! Sincerely, Chen
Re: [ccp4bb] refine structure with mtz file in P212121 or P222?
Point 1) You don't need to reindex the data to change SG from P222 to P212121. Indexing and integration use the lattice type , and scaling the point group - not the space group. The only difference between the 2 space groups is that the odd axial reflections will be assent in P212121. Point 2) That is very very strange - if your solution is found for SG P212121 it is surprising if the refinement works better in P222! But check how the refinement and map generation is done. If the PDB file defines the SG as P212121 then maybe the refinement assumes the mtz file should also have that symmetry.. On 5 September 2014 16:54, Wei Shi wei.shi...@gmail.com wrote: Hi all, I am working with a protein-ligand complex structure. The data was indexed with XDS and mtz file is generated using Phenix---Reflection tools---Reflection file editor. The space group used was P222. Then, I used the Phenix---Molecular replacement to find a solution (try all possible in same pointgroup), and the MR solution was in space group P212121. And then, I reindexed the data in XDS using space group P212121 and then generate the mtz file in space group P212121. So, I have two mtz file, in space group P212121 and P222. When I refine the MR solution with the different mtz files, the map generated is not same for the ligand. The statisitcs is better with P222 mtz file than P212121 mtz file. I am wondering why the density for the ligand and the statistics is different using mtz file in different space groups and which mtz file should I use to get the final structure model. Thank you so much! Best, Wei
Re: [ccp4bb] Extract Euler angles from fractional coordinate matrix
I am sorry for my carelessness on the definition of Euler angles. I am just thinking of an Euler angle equivalent. Sorry for the confusion I have made. On Fri, Sep 5, 2014 at 3:34 PM, Eleanor Dodson eleanor.dod...@york.ac.uk wrote: I don't think Eulerian angles are defined for a non-orthogonal axis system? ? They are defined relative to perpendicular axes X Y Z e.g. Rotate coordinates by gamma about Z, beta about Y', alpha about Z. Eleanor On 5 September 2014 16:27, Chen Zhao c.z...@yale.edu wrote: Thank you Eleanor for your reply. I am actually considering how to describe a pseudo-NCS with an arbitrary rotational and translational relationship. I don't have to do this but I am just curious. It is more straightforward if I say how the two molecules are related by a rotation around unit cell axis than around orthogonal coordinate axis, which does not have an absolute physical meaning. The command line output after coot superpose prints out the rotational and translational matrices for both the orthogonal and fractional coordinate system. For using coordconv, my concern is that if I deal with a low-symmetry unit cell, which is not orthogonal by itself, the Euler angles for the fractional coordinate system and the orthogonal coordinate system should be different. If I just feed some numbers into coordconv, will it consider them as orthogonal coordinates? Thank you, Chen On Fri, Sep 5, 2014 at 6:24 AM, Eleanor Dodson eleanor.dod...@york.ac.uk wrote: Rotation matrices are rarely specified in a fractional coordinate system? The criteria for checking such a matrix is Is the determinant 1? and this only holds for orthogonal matrices. I guess the way I would do this though. You presumably have two sets of fractional coordinates, before and after rotation? There is a ccp4 program - coordconv which will read the fractional coordinates and generate pdb format with the convention ncode = 1 (You may need to fudge the fractional format I suppose..) You can then use superpose to match the two sets of coordinates and the output will tell you the Eulerian angles used for the rotation! Lots of ways to kill cats! Eleanor On 4 September 2014 21:21, Phil Jeffrey pjeff...@princeton.edu wrote: The orthogonal/fractional matrix is outlined here: http://www.iucr.org/__data/assets/pdf_file/0009/7011/19_ 06_cowtan_coordinate_frames.pdf Sorry to say I apparently ditched my old Fortran o2f and f2o programs to do that. Bear in mind, however, that orthogonal has no fixed orientation with respect to fractional - for most space groups ncode 1 is often used but for primitive monoclinic ncode 3 is sometimes used, and I think the matrix shown in Kevin Cowtan's document above corresponds to ncode 1. Phil Jeffrey Princeton On 9/4/14 3:55 PM, Chen Zhao wrote: I am sorry, just to clarify, the fractional coordinate matrix I referred to is a rotational matrix in the fractional coordinate system. On Thu, Sep 4, 2014 at 3:52 PM, Chen Zhao c.z...@yale.edu mailto:c.z...@yale.edu wrote: Hi all, I am just curious whether there are some tools extracting the Euler angles from a fractional coordinate matrix. I have no luck searching it online. Alternatively, I found the analytical solution for the Euler angles from an orthogonal coordinate matrix. So in the worst case, my problem reduces to calculating the transformation matrix between the fractional and orthogonal coordinate system. I feel a little bit at a loss because it is 6 years since I last studied linear algebra. How can I calculate this for a specific unit cell? Thanks a lot in advance! Sincerely, Chen
Re: [ccp4bb] refine structure with mtz file in P212121 or P222?
As always, look at the unit cell packing of your alternative solutions. In all likelihood one of these two solutions from Phaser should pack sensibly in the unit cell, and the other will not. You may get some sort of quasi-reasonable-looking electron density out of the wrong solution initially, but usually not. It is possible that you indexed, integrated, and scaled in P222 and P212121, but if you do a full automated search in Phaser it will look for solutions in both space groups, and may have written out a solution in the same spacegroup both times. (Check the log file or PDB header.) Cheers, ___ Roger S. Rowlett Gordon Dorothy Kline Professor Department of Chemistry Colgate University 13 Oak Drive Hamilton, NY 13346 tel: (315)-228-7245 ofc: (315)-228-7395 fax: (315)-228-7935 email: rrowl...@colgate.edu On 9/5/2014 11:54 AM, Wei Shi wrote: Hi all, I am working with a protein-ligand complex structure. The data was indexed with XDS and mtz file is generated using Phenix---Reflection tools---Reflection file editor. The space group used was P222. Then, I used the Phenix---Molecular replacement to find a solution (try all possible in same pointgroup), and the MR solution was in space group P212121. And then, I reindexed the data in XDS using space group P212121 and then generate the mtz file in space group P212121. So, I have two mtz file, in space group P212121 and P222. When I refine the MR solution with the different mtz files, the map generated is not same for the ligand. The statisitcs is better with P222 mtz file than P212121 mtz file. I am wondering why the density for the ligand and the statistics is different using mtz file in different space groups and which mtz file should I use to get the final structure model. Thank you so much! Best, Wei
Re: [ccp4bb] refine structure with mtz file in P212121 or P222?
Thank you! Yes, the PDB file defines the SG as P212121. When I load pdb file from phaser (P222 mtz file as data and the solution is in P212121) into Phenix.refine, Phenix.refine automatically assumes that the space group is P212121. I am not sure in this case why using P222 mtz and P212121 mtz to do refinement would make a difference in the map. The difference is not very big but it's different for some atoms of the ligand Thank you! Best, Wei On Fri, Sep 5, 2014 at 3:40 PM, Eleanor Dodson eleanor.dod...@york.ac.uk wrote: Point 1) You don't need to reindex the data to change SG from P222 to P212121. Indexing and integration use the lattice type , and scaling the point group - not the space group. The only difference between the 2 space groups is that the odd axial reflections will be assent in P212121. Point 2) That is very very strange - if your solution is found for SG P212121 it is surprising if the refinement works better in P222! But check how the refinement and map generation is done. If the PDB file defines the SG as P212121 then maybe the refinement assumes the mtz file should also have that symmetry.. On 5 September 2014 16:54, Wei Shi wei.shi...@gmail.com wrote: Hi all, I am working with a protein-ligand complex structure. The data was indexed with XDS and mtz file is generated using Phenix---Reflection tools---Reflection file editor. The space group used was P222. Then, I used the Phenix---Molecular replacement to find a solution (try all possible in same pointgroup), and the MR solution was in space group P212121. And then, I reindexed the data in XDS using space group P212121 and then generate the mtz file in space group P212121. So, I have two mtz file, in space group P212121 and P222. When I refine the MR solution with the different mtz files, the map generated is not same for the ligand. The statisitcs is better with P222 mtz file than P212121 mtz file. I am wondering why the density for the ligand and the statistics is different using mtz file in different space groups and which mtz file should I use to get the final structure model. Thank you so much! Best, Wei