Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Hi, I agree with Melanie and I think we should also go further... We could require reviewers to tick a box on a PDB reviewer page to confirm they have “checked” the structure. Whether this box has been ticked could then be displayed on the PDB page (of course the reviewer would remain anonymous) or sent to the journal that published the paper (which could them prompt the reviewer to do so). It would be transparent for both depositors, editors, and users whether the structure had been reviewed. In papers with multiple reviewers, of course not all reviewers would be expected to look (i.e. interdisciplinary papers). As John says, many journals already require validation reports — this is only one small step further and wouldn’t unduly increase the work load for reviewers, since we already expect them to check the structure. I’m not saying it would completely solve the problem as people could just log on and click the button without reading, but it would at least be the start of accountability or at least record keeping. If you forced (assigned) reviewers to visit the PDB webpage, seeing a large array of red sliders might at least trigger a reviewer to look more carefully. Even better, a potential reviewer page could contain automatically generated electron density pictures for ligands (as are already available on the pdb websites, so generating them isn’t extra work) — this would avoid the problem where images in the paper are inaccurately or incorrectly generated. Again, it might trigger reviewer investigation if they’re forced to click a button on the websites for “I am ok with this structure and it is of appropriate quality for the conclusions of the paper”. This would of course require journals to require links to the pdb depositions on manuscript submission rather than asking for them if a reviewer requests it — I don’t think that’s a big ask. Thanks, Nick > On 23 Jul 2019, at 09:52, "melanie.voll...@diamond.ac.uk" > > Dear John, > > > Yes, I think the PDB should be stricter. The PDB is in the position to > enforce compliance with rules and if they are not followed then one doesn't > get a validation report and in turn it would be difficult to publish (most > journals require a validation report). For years it has been tried to make > adherence to standards voluntary but there are still examples, like the one > that started the discussion, where people just don't bother. > > > For these particular structures I actually suspect that even if there were > all reports given to the referees they may not have looked carefully enough. > The journal where they were published in is heavy on the chemistry side so > perhaps the referees focused on the synthesis of the compounds rather than > the structure and the interaction within the protein. Aside from that, at a > resolution of 2.7A to 3A discussing detailed chemical interactions and > placing a compound with confidence is questionable anyway. > > > Cheers > > > M > > > From: CCP4 bulletin board on behalf of John > Berrisford > Sent: 22 July 2019 22:23:47 > To: ccp4bb > Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 > > Dear Harry > > We will be shortly making it mandatory for depositors to provide a value for > at least one of the merging statistics (Rmerge, Rpim, CC1/2 etc..). Most > depositors do, but we want to ensure that all depositors do provide at least > one value for a merging metric. > > We would welcome feedback if we should be stricter and require a (or more > than one) specific metric (e.g. CC1/2) – please be aware that any required > metric must be available from all merging/scaling software. > > Thanks > > John > > From: CCP4BB > Sent: 22 July 2019 11:32 > To: John Berrisford > Cc: CCP4BB@JISCMAIL.AC.UK > Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 > > Hi John > > These are great, but the things that make me suspicious are the values of > overall R(merge); these are tucked away in the full reports, rather than > highlighted with all the other structural metrics in the validation sliders. > It would be wonderful to be able to see at a glance where overall R(merge) > values like these fit in with those of other deposited structures (even > better if it could be drawn to the attention of authors before final > deposition). > > Kay and Gérard have already pointed out that the data processing here may > have some issues. > > Of course, those of us involved in teaching data processing have been > emphasizing the importance of CC(1/2) rather than relying on R(merge) for > yonks, but if CC(1/2) isn't given in the report it's all we have to go &g
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Dear John, Yes, I think the PDB should be stricter. The PDB is in the position to enforce compliance with rules and if they are not followed then one doesn't get a validation report and in turn it would be difficult to publish (most journals require a validation report). For years it has been tried to make adherence to standards voluntary but there are still examples, like the one that started the discussion, where people just don't bother. For these particular structures I actually suspect that even if there were all reports given to the referees they may not have looked carefully enough. The journal where they were published in is heavy on the chemistry side so perhaps the referees focused on the synthesis of the compounds rather than the structure and the interaction within the protein. Aside from that, at a resolution of 2.7A to 3A discussing detailed chemical interactions and placing a compound with confidence is questionable anyway. Cheers M From: CCP4 bulletin board on behalf of John Berrisford Sent: 22 July 2019 22:23:47 To: ccp4bb Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Dear Harry We will be shortly making it mandatory for depositors to provide a value for at least one of the merging statistics (Rmerge, Rpim, CC1/2 etc..). Most depositors do, but we want to ensure that all depositors do provide at least one value for a merging metric. We would welcome feedback if we should be stricter and require a (or more than one) specific metric (e.g. CC1/2) – please be aware that any required metric must be available from all merging/scaling software. Thanks John From: CCP4BB Sent: 22 July 2019 11:32 To: John Berrisford Cc: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Hi John These are great, but the things that make me suspicious are the values of overall R(merge); these are tucked away in the full reports, rather than highlighted with all the other structural metrics in the validation sliders. It would be wonderful to be able to see at a glance where overall R(merge) values like these fit in with those of other deposited structures (even better if it could be drawn to the attention of authors before final deposition). Kay and Gérard have already pointed out that the data processing here may have some issues. Of course, those of us involved in teaching data processing have been emphasizing the importance of CC(1/2) rather than relying on R(merge) for yonks, but if CC(1/2) isn't given in the report it's all we have to go Harry -- Dr Harry Powell On 22 Jul 2019, at 10:05, John Berrisford mailto:j...@ebi.ac.uk>> wrote: Dear Daniel The issues you mentioned are highlighted in the wwPDB validation report http://www.ebi.ac.uk/pdbe/entry-files/6mo0_full_validation.pdf and global issues with the entry are highlighted in the validation sliders http://www.ebi.ac.uk/pdbe/entry/pdb/6MO0 The validation sliders are shown on the entry pages at all wwPDB PDB sites (RCSB, PDBe and PDBj) and they all provide a link to download the wwPDB validation report directly from the entry page. Regards John From: CCP4 bulletin board mailto:CCP4BB@JISCMAIL.AC.UK>> On Behalf Of Bonsor, Daniel Sent: 19 July 2019 18:43 To: CCP4BB@JISCMAIL.AC.UK<mailto:CCP4BB@JISCMAIL.AC.UK> Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Would it be possible to add a public annotations section to the PDB, to allow us to potentially flag/warn whoever downloads that particular structure, there could be something wrong with it, such as wrong space group, no/poor density fitting for ligand. Something similar to PubPeer maybe? Daniel A. Bonsor PhD Institute of Human Virology, University of Maryland at Baltimore 725 W. Lombard Street N571 Baltimore MD 21201 From: CCP4 bulletin board mailto:CCP4BB@JISCMAIL.AC.UK>> on behalf of Patrick Loll mailto:pjl...@gmail.com>> Sent: Friday, July 19, 2019 1:17 PM To: CCP4BB@JISCMAIL.AC.UK<mailto:CCP4BB@JISCMAIL.AC.UK> mailto:CCP4BB@JISCMAIL.AC.UK>> Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 The idea of contacting the editor (and/or author) is an excellent one, and indeed the correct thing to do scientifically. However, I’m disillusioned: I’ve been down this path before with a high-profile vanity journal, and while the editors paid lip service to the notion that the record should be corrected, in reality they led me on for the better part of a year, and got me to write up detailed analyses of why the ligand positioning was not justified, before eventually saying “no, we don’t see any need to publish a correction.” I speculate that the journal prefers not avoid corrections, for fear that too many corrections will make the journal a less desirable destination. > On 19 Jul 2019, at 11:23 AM, Bärbel Blaum
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Following up what Kay Diederichs recently wrote, each page of Proteopedia that contains user added comments have a visible banner on the top, e.g. see: https://proteopedia.org/w/2bs2 n.b. these user added comments are all signed, i.e. they are NOT anonymous, so it is a good location to consider user discussion on structural studies. Joel On 22Jul, 2019, at 14:16, Kay Diederichs mailto:kay.diederi...@uni-konstanz.de>> wrote: Proteopedia has this - see https://proteopedia.org/wiki/index.php/Proteopedia:Comments_on_Published_Structures Kay On Fri, 19 Jul 2019 17:42:33 +, Bonsor, Daniel mailto:dbon...@som.umaryland.edu>> wrote: Would it be possible to add a public annotations section to the PDB, to allow us to potentially flag/warn whoever downloads that particular structure, there could be something wrong with it, such as wrong space group, no/poor density fitting for ligand. Something similar to PubPeer maybe? Daniel A. Bonsor PhD Institute of Human Virology, University of Maryland at Baltimore 725 W. Lombard Street N571 Baltimore MD 21201 From: CCP4 bulletin board mailto:CCP4BB@JISCMAIL.AC.UK>> on behalf of Patrick Loll mailto:pjl...@gmail.com>> Sent: Friday, July 19, 2019 1:17 PM To: CCP4BB@JISCMAIL.AC.UK<mailto:CCP4BB@JISCMAIL.AC.UK> mailto:CCP4BB@JISCMAIL.AC.UK>> Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 The idea of contacting the editor (and/or author) is an excellent one, and indeed the correct thing to do scientifically. However, I�m disillusioned: I�ve been down this path before with a high-profile vanity journal, and while the editors paid lip service to the notion that the record should be corrected, in reality they led me on for the better part of a year, and got me to write up detailed analyses of why the ligand positioning was not justified, before eventually saying �no, we don�t see any need to publish a correction.� I speculate that the journal prefers not avoid corrections, for fear that too many corrections will make the journal a less desirable destination. On 19 Jul 2019, at 11:23 AM, B�rbel Blaum mailto:baerbel.bl...@intherabio.com>> wrote: Hi Rhys, the reported B-factors for the �ligands� are all way below the reported B-factors for the protein chains, with the worst of the three complexes reporting unitless numbers 23.2 and 64.8, respectively, just to highlight *one* striking feature of the data collection and refinement table. So even with the limited info normally available to reviewers this table should have raised a red flag. After the re-refinement suggested by others, i.e. your own proper assessment of the crystallographic data, if you do not find noteworthy density you may want to contact the article�s editor with your results. If you work in this field, i.e. really care about this paper scientifically and you are not afraid to confront the authors you could suggest writing a comment/direct response article but in my opinion that would only make sense if you can be sure beforehand that it will be linked visibly to the actual paper, else it will be a waste of time. And don�t forget that just because one or some of the authors did a bad job at the crystallographic end other findings of the paper might still be solid � in collaborations often one author is unable to critically evaluate another author�s contribution and this would not be the first case were good synthetic or biological work is presented along with a bad crystal structure. By the way and a bit ironically this protein may have suffered bad crystallography/scientific practice before - I think it was one of the fake Krishna Murthy structures, right? The associated (now retracted) article I mean is here https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub [https://ars.els-cdn.com/content/image/S00222836.gif]<https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub> RETRACTED: Crystal structure of dengue virus NS3 protease in complex with a bowman-birk inhibitor: implications for flaviviral polyprotein processing and drug design - ScienceDirect - ScienceDirect.com<http://ScienceDirect.com> | Science, health and medical journals, full text articles and books.<https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub> www.sciencedirect.com<http://www.sciencedirect.com> COMMUNIC Crystal Structure of Dengue Complex with a Bowman-Bir ro L. 1Center for Macromolecular C f A 8 U T MCLM 244, Birmingham AL 35294-0005, USA 2Department of Biochemistry and Molecular Biology, Kansas University Medical Center 3901 Rainbow Boulevard Kansas City, KS 66160- 7421, USA Dengue viruses are members of the Flaviviridae and cause dengue fever Dengue fever and dengue hemorrhagic ... Kind regards, B�rbel --- B�rbel Blaum, PhD Inthera Bioscience AG Einsiedlerstrasse 34 CH-8820 Waedenswil Swi
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Dear Harry We will be shortly making it mandatory for depositors to provide a value for at least one of the merging statistics (Rmerge, Rpim, CC1/2 etc..). Most depositors do, but we want to ensure that all depositors do provide at least one value for a merging metric. We would welcome feedback if we should be stricter and require a (or more than one) specific metric (e.g. CC1/2) – please be aware that any required metric must be available from all merging/scaling software. Thanks John From: CCP4BB Sent: 22 July 2019 11:32 To: John Berrisford Cc: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Hi John These are great, but the things that make me suspicious are the values of overall R(merge); these are tucked away in the full reports, rather than highlighted with all the other structural metrics in the validation sliders. It would be wonderful to be able to see at a glance where overall R(merge) values like these fit in with those of other deposited structures (even better if it could be drawn to the attention of authors before final deposition). Kay and Gérard have already pointed out that the data processing here may have some issues. Of course, those of us involved in teaching data processing have been emphasizing the importance of CC(1/2) rather than relying on R(merge) for yonks, but if CC(1/2) isn't given in the report it's all we have to go Harry -- Dr Harry Powell On 22 Jul 2019, at 10:05, John Berrisford mailto:j...@ebi.ac.uk> > wrote: Dear Daniel The issues you mentioned are highlighted in the wwPDB validation report http://www.ebi.ac.uk/pdbe/entry-files/6mo0_full_validation.pdf and global issues with the entry are highlighted in the validation sliders http://www.ebi.ac.uk/pdbe/entry/pdb/6MO0 The validation sliders are shown on the entry pages at all wwPDB PDB sites (RCSB, PDBe and PDBj) and they all provide a link to download the wwPDB validation report directly from the entry page. Regards John From: CCP4 bulletin board mailto:CCP4BB@JISCMAIL.AC.UK> > On Behalf Of Bonsor, Daniel Sent: 19 July 2019 18:43 To: CCP4BB@JISCMAIL.AC.UK <mailto:CCP4BB@JISCMAIL.AC.UK> Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Would it be possible to add a public annotations section to the PDB, to allow us to potentially flag/warn whoever downloads that particular structure, there could be something wrong with it, such as wrong space group, no/poor density fitting for ligand. Something similar to PubPeer maybe? Daniel A. Bonsor PhD Institute of Human Virology, University of Maryland at Baltimore 725 W. Lombard Street N571 Baltimore MD 21201 _ From: CCP4 bulletin board mailto:CCP4BB@JISCMAIL.AC.UK> > on behalf of Patrick Loll mailto:pjl...@gmail.com> > Sent: Friday, July 19, 2019 1:17 PM To: CCP4BB@JISCMAIL.AC.UK <mailto:CCP4BB@JISCMAIL.AC.UK> mailto:CCP4BB@JISCMAIL.AC.UK> > Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 The idea of contacting the editor (and/or author) is an excellent one, and indeed the correct thing to do scientifically. However, I’m disillusioned: I’ve been down this path before with a high-profile vanity journal, and while the editors paid lip service to the notion that the record should be corrected, in reality they led me on for the better part of a year, and got me to write up detailed analyses of why the ligand positioning was not justified, before eventually saying “no, we don’t see any need to publish a correction.” I speculate that the journal prefers not avoid corrections, for fear that too many corrections will make the journal a less desirable destination. > On 19 Jul 2019, at 11:23 AM, Bärbel Blaum <mailto:baerbel.bl...@intherabio.com> > wrote: > > Hi Rhys, > the reported B-factors for the “ligands” are all way below the reported > B-factors for the protein chains, with the worst of the three complexes > reporting unitless numbers 23.2 and 64.8, respectively, just to highlight > *one* striking feature of the data collection and refinement table. So even > with the limited info normally available to reviewers this table should have > raised a red flag. After the re-refinement suggested by others, i.e. your own > proper assessment of the crystallographic data, if you do not find noteworthy > density you may want to contact the article’s editor with your results. If > you work in this field, i.e. really care about this paper scientifically and > you are not afraid to confront the authors you could suggest writing a > comment/direct response article but in my opinion that would only make sense > if you can be sure beforehand that it will be linked visibly to the actual > paper, else it will be a waste of time. And don’t forget that just because > on
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Well it is also on the webpage of the entry but I suppose not all eight authors listed in this field did the crystallography. Since the pdb has not come up with a set of “contribution” fields to be filled in yet (maybe that would actually be useful!) the person in charge of the actual submission can file the full authors list as deposition author if (s)he likes and it seems indeed quite common that the full list of publication authors ends up as “deposition authors”. Which is not surprising if you consider in which way many depositions are being done – by a first-time PhD student who is totally overwhelmed when he realizes that the pdb deposition takes approximately five times as long as (s)he expected. -- Bärbel Blaum, PhD Inthera Bioscience AG Einsiedlerstrasse 34 CH-8820 Waedenswil Switzerland E-Mail: baerbel.bl...@intherabio.com Phone: +41 43 477 94 72-- Von: CCP4 bulletin board im Auftrag von David Schuller Organisation: Cornell University Antworten an: Datum: Montag, 22. Juli 2019 um 14:45 An: Betreff: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 On 7/22/19 3:40 AM, Bärbel Blaum wrote: "As I said, the lead author may not actually be the crystallographer here" Bärbel No need to wonder, that info should be in the PDB file header. The author of the coordinates is not necessarily the same as the author of the primary reference. -- === All Things Serve the Beam === David J. Schuller modern man in a post-modern world MacCHESS, Cornell University schul...@cornell.edu To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1 To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
On 7/22/19 3:40 AM, Bärbel Blaum wrote: "As I said, the lead author may not actually be the crystallographer here" Bärbel No need to wonder, that info should be in the PDB file header. The author of the coordinates is not necessarily the same as the author of the primary reference. -- === All Things Serve the Beam === David J. Schuller modern man in a post-modern world MacCHESS, Cornell University schul...@cornell.edu To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Proteopedia has this - see https://proteopedia.org/wiki/index.php/Proteopedia:Comments_on_Published_Structures Kay On Fri, 19 Jul 2019 17:42:33 +, Bonsor, Daniel wrote: >Would it be possible to add a public annotations section to the PDB, to allow >us to potentially flag/warn whoever downloads that particular structure, there >could be something wrong with it, such as wrong space group, no/poor density >fitting for ligand. Something similar to PubPeer maybe? > > >Daniel A. Bonsor PhD >Institute of Human Virology, >University of Maryland at Baltimore >725 W. Lombard Street N571 >Baltimore >MD 21201 > > > >From: CCP4 bulletin board on behalf of Patrick Loll > >Sent: Friday, July 19, 2019 1:17 PM >To: CCP4BB@JISCMAIL.AC.UK >Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 > >The idea of contacting the editor (and/or author) is an excellent one, and >indeed the correct thing to do scientifically. However, I�m disillusioned: >I�ve been down this path before with a high-profile vanity journal, and while >the editors paid lip service to the notion that the record should be >corrected, in reality they led me on for the better part of a year, and got me >to write up detailed analyses of why the ligand positioning was not justified, >before eventually saying �no, we don�t see any need to publish a correction.� >I speculate that the journal prefers not avoid corrections, for fear that too >many corrections will make the journal a less desirable destination. > >> On 19 Jul 2019, at 11:23 AM, B�rbel Blaum >> wrote: >> >> Hi Rhys, >> the reported B-factors for the �ligands� are all way below the reported >> B-factors for the protein chains, with the worst of the three complexes >> reporting unitless numbers 23.2 and 64.8, respectively, just to highlight >> *one* striking feature of the data collection and refinement table. So even >> with the limited info normally available to reviewers this table should have >> raised a red flag. After the re-refinement suggested by others, i.e. your >> own proper assessment of the crystallographic data, if you do not find >> noteworthy density you may want to contact the article�s editor with your >> results. If you work in this field, i.e. really care about this paper >> scientifically and you are not afraid to confront the authors you could >> suggest writing a comment/direct response article but in my opinion that >> would only make sense if you can be sure beforehand that it will be linked >> visibly to the actual paper, else it will be a waste of time. And don�t >> forget that just because one or some of the authors did a bad job at the >> crystallographic end other findings of the paper might still be solid � in >> collaborations often one author is unable to critically evaluate another >> author�s contribution and this would not be the first case were good >> synthetic or biological work is presented along with a bad crystal structure. >> By the way and a bit ironically this protein may have suffered bad >> crystallography/scientific practice before - I think it was one of the fake >> Krishna Murthy structures, right? The associated (now retracted) article I >> mean is here >> https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub >[https://ars.els-cdn.com/content/image/S00222836.gif]<https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub> >RETRACTED: Crystal structure of dengue virus NS3 protease in complex with a >bowman-birk inhibitor: implications for flaviviral polyprotein processing and >drug design - ScienceDirect - ScienceDirect.com | Science, health and medical >journals, full text articles and >books.<https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub> >www.sciencedirect.com >COMMUNIC Crystal Structure of Dengue Complex with a Bowman-Bir ro L. 1Center >for Macromolecular C f A 8 U T MCLM 244, Birmingham AL 35294-0005, USA >2Department of Biochemistry and Molecular Biology, Kansas University Medical >Center 3901 Rainbow Boulevard Kansas City, KS 66160- 7421, USA Dengue viruses >are members of the Flaviviridae and cause dengue fever Dengue fever and dengue >hemorrhagic ... > >> Kind regards, B�rbel >> --- >> B�rbel Blaum, PhD >> Inthera Bioscience AG >> Einsiedlerstrasse 34 >> CH-8820 Waedenswil >> Switzerland >> E-Mail: baerbel.bl...@intherabio.com >> Phone: +41 43 477 94 72-- >> >> >> >> Von: CCP4 bulletin board im Auftrag von "Manfred S. >> Weiss" >> Antworten an: "Manfred S
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Hi John These are great, but the things that make me suspicious are the values of overall R(merge); these are tucked away in the full reports, rather than highlighted with all the other structural metrics in the validation sliders. It would be wonderful to be able to see at a glance where overall R(merge) values like these fit in with those of other deposited structures (even better if it could be drawn to the attention of authors before final deposition). Kay and Gérard have already pointed out that the data processing here may have some issues. Of course, those of us involved in teaching data processing have been emphasizing the importance of CC(1/2) rather than relying on R(merge) for yonks, but if CC(1/2) isn't given in the report it's all we have to go Harry -- Dr Harry Powell > On 22 Jul 2019, at 10:05, John Berrisford wrote: > > Dear Daniel > > The issues you mentioned are highlighted in the wwPDB validation report > http://www.ebi.ac.uk/pdbe/entry-files/6mo0_full_validation.pdf > and global issues with the entry are highlighted in the validation sliders > http://www.ebi.ac.uk/pdbe/entry/pdb/6MO0 > > The validation sliders are shown on the entry pages at all wwPDB PDB sites > (RCSB, PDBe and PDBj) and they all provide a link to download the wwPDB > validation report directly from the entry page. > > Regards > > John > > From: CCP4 bulletin board On Behalf Of Bonsor, Daniel > Sent: 19 July 2019 18:43 > To: CCP4BB@JISCMAIL.AC.UK > Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 > > Would it be possible to add a public annotations section to the PDB, to allow > us to potentially flag/warn whoever downloads that particular structure, > there could be something wrong with it, such as wrong space group, no/poor > density fitting for ligand. Something similar to PubPeer maybe? > > > > Daniel A. Bonsor PhD > Institute of Human Virology, > University of Maryland at Baltimore > 725 W. Lombard Street N571 > Baltimore > MD 21201 > > > From: CCP4 bulletin board on behalf of Patrick Loll > > Sent: Friday, July 19, 2019 1:17 PM > To: CCP4BB@JISCMAIL.AC.UK > Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 > > The idea of contacting the editor (and/or author) is an excellent one, and > indeed the correct thing to do scientifically. However, I’m disillusioned: > I’ve been down this path before with a high-profile vanity journal, and while > the editors paid lip service to the notion that the record should be > corrected, in reality they led me on for the better part of a year, and got > me to write up detailed analyses of why the ligand positioning was not > justified, before eventually saying “no, we don’t see any need to publish a > correction.” I speculate that the journal prefers not avoid corrections, for > fear that too many corrections will make the journal a less desirable > destination. > > > On 19 Jul 2019, at 11:23 AM, Bärbel Blaum > > wrote: > > > > Hi Rhys, > > the reported B-factors for the “ligands” are all way below the reported > > B-factors for the protein chains, with the worst of the three complexes > > reporting unitless numbers 23.2 and 64.8, respectively, just to highlight > > *one* striking feature of the data collection and refinement table. So even > > with the limited info normally available to reviewers this table should > > have raised a red flag. After the re-refinement suggested by others, i.e. > > your own proper assessment of the crystallographic data, if you do not find > > noteworthy density you may want to contact the article’s editor with your > > results. If you work in this field, i.e. really care about this paper > > scientifically and you are not afraid to confront the authors you could > > suggest writing a comment/direct response article but in my opinion that > > would only make sense if you can be sure beforehand that it will be linked > > visibly to the actual paper, else it will be a waste of time. And don’t > > forget that just because one or some of the authors did a bad job at the > > crystallographic end other findings of the paper might still be solid – in > > collaborations often one author is unable to critically evaluate another > > author’s contribution and this would not be the first case were good > > synthetic or biological work is presented along with a bad crystal > > structure. > > By the way and a bit ironically this protein may have suffered bad > > crystallography/scientific practice before - I think it was one of the fake > > Krishna Murthy structures, right? The associated (now retracted) article I > &
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Dear Daniel The issues you mentioned are highlighted in the wwPDB validation report http://www.ebi.ac.uk/pdbe/entry-files/6mo0_full_validation.pdf and global issues with the entry are highlighted in the validation sliders http://www.ebi.ac.uk/pdbe/entry/pdb/6MO0 The validation sliders are shown on the entry pages at all wwPDB PDB sites (RCSB, PDBe and PDBj) and they all provide a link to download the wwPDB validation report directly from the entry page. Regards John From: CCP4 bulletin board On Behalf Of Bonsor, Daniel Sent: 19 July 2019 18:43 To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Would it be possible to add a public annotations section to the PDB, to allow us to potentially flag/warn whoever downloads that particular structure, there could be something wrong with it, such as wrong space group, no/poor density fitting for ligand. Something similar to PubPeer maybe? Daniel A. Bonsor PhD Institute of Human Virology, University of Maryland at Baltimore 725 W. Lombard Street N571 Baltimore MD 21201 _ From: CCP4 bulletin board mailto:CCP4BB@JISCMAIL.AC.UK> > on behalf of Patrick Loll mailto:pjl...@gmail.com> > Sent: Friday, July 19, 2019 1:17 PM To: CCP4BB@JISCMAIL.AC.UK <mailto:CCP4BB@JISCMAIL.AC.UK> mailto:CCP4BB@JISCMAIL.AC.UK> > Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 The idea of contacting the editor (and/or author) is an excellent one, and indeed the correct thing to do scientifically. However, Im disillusioned: Ive been down this path before with a high-profile vanity journal, and while the editors paid lip service to the notion that the record should be corrected, in reality they led me on for the better part of a year, and got me to write up detailed analyses of why the ligand positioning was not justified, before eventually saying no, we dont see any need to publish a correction. I speculate that the journal prefers not avoid corrections, for fear that too many corrections will make the journal a less desirable destination. > On 19 Jul 2019, at 11:23 AM, Bärbel Blaum mailto:baerbel.bl...@intherabio.com> > wrote: > > Hi Rhys, > the reported B-factors for the ligands are all way below the reported B-factors for the protein chains, with the worst of the three complexes reporting unitless numbers 23.2 and 64.8, respectively, just to highlight *one* striking feature of the data collection and refinement table. So even with the limited info normally available to reviewers this table should have raised a red flag. After the re-refinement suggested by others, i.e. your own proper assessment of the crystallographic data, if you do not find noteworthy density you may want to contact the articles editor with your results. If you work in this field, i.e. really care about this paper scientifically and you are not afraid to confront the authors you could suggest writing a comment/direct response article but in my opinion that would only make sense if you can be sure beforehand that it will be linked visibly to the actual paper, else it will be a waste of time. And dont forget that just because one or some of the authors did a bad job at the crystallographic end other findings of the paper might still be solid in collaborations often one author is unable to critically evaluate another authors contribution and this would not be the first case were good synthetic or biological work is presented along with a bad crystal structure. > By the way and a bit ironically this protein may have suffered bad crystallography/scientific practice before - I think it was one of the fake Krishna Murthy structures, right? The associated (now retracted) article I mean is here > https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dih ub <https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Di hub> <https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Di hub> RETRACTED: Crystal structure of dengue virus NS3 protease in complex with a bowman-birk inhibitor: implications for flaviviral polyprotein processing and drug design - ScienceDirect - ScienceDirect.com | Science, health and medical journals, full text articles and books. www.sciencedirect.com <http://www.sciencedirect.com> COMMUNIC Crystal Structure of Dengue Complex with a Bowman-Bir ro L. 1Center for Macromolecular C f A 8 U T MCLM 244, Birmingham AL 35294-0005, USA 2Department of Biochemistry and Molecular Biology, Kansas University Medical Center 3901 Rainbow Boulevard Kansas City, KS 66160- 7421, USA Dengue viruses are members of the Flaviviridae and cause dengue fever Dengue fever and dengue hemorrhagic ... > Kind regards, Bärbel > --- > Bärbel Blaum, PhD > Inthera Bioscience AG > Einsiedlerstrasse 34 > CH-8820 Waedenswil > Switzerland > E-Mail: bae
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
You are absolutely right of course, Jacob! I just did that “just for fun”. As I said, the lead author may not actually be the crystallographer here and if that is the case it is possible that he would actually appreciate some feedback from this list. Despite the bad experiences some of us had with editors from glossy journals we should not assume bad intentions by all the authors as a group. So let’s see what he says. Bärbel -- Bärbel Blaum, PhD Inthera Bioscience AG Einsiedlerstrasse 34 CH-8820 Waedenswil Switzerland E-Mail: baerbel.bl...@intherabio.com Phone: +41 43 477 94 72-- Von: CCP4 bulletin board im Auftrag von "Keller, Jacob" Antworten an: "Keller, Jacob" Datum: Freitag, 19. Juli 2019 um 23:09 An: Betreff: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 I don’t think anyone mentioned contacting the authors first—doesn’t this seem like the first thing one should do? Jacob + Jacob Pearson Keller Research Scientist / Looger Lab HHMI Janelia Research Campus 19700 Helix Dr, Ashburn, VA 20147 Desk: (571)209-4000 x3159 Cell: (301)592-7004 + The content of this email is confidential and intended for the recipient specified in message only. It is strictly forbidden to share any part of this message with any third party, without a written consent of the sender. If you received this message by mistake, please reply to this message and follow with its deletion, so that we can ensure such a mistake does not occur in the future. From: CCP4 bulletin board On Behalf Of Roger Rowlett Sent: Friday, July 19, 2019 4:48 PM To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 If it is as it appears, it is disappointing to see this in JACS. I would expect better. Unfortunately, reviewers don't always get a lot of information to judge quality of structures (which has been discussed extensively on prior occasions on this board), so some trust is involved that what data is presented by the authors is representative. I see the authors duly presented a 2Fo-Fc map and a Fo-Fc omit map in one of the figures, but it appears from a personal examination that the Fo-Fc omit map presented may not resemble what one gets when personally examining the deposited data, and some protein chain refinement problems are immediately evident. That is a bit concerning. High-pressure research, coupled with (perhaps) a lack of understanding, experience, and supervision is a potentially dangerous ethical stew. Correcting erroneous published data is challenging. I share the frustration of others when attempting to challenge scientifically questionable published results. My trip down the rabbit hole (not related directly to crystallography) nearly identical to that related previously here. Spend your own time writing a full paper rebuttal (ugh) or move on and concentrate resources on your own work. _ Roger S. Rowlett Gordon & Dorothy Kline Professor, Emeritus Department of Chemistry Colgate University email: rrowl...@colgate.edu On 7/19/2019 2:24 PM, Robbie Joosten wrote: Even though PDB-REDO cannot salvage this model without extensive rebuilding which is what Tristan showed, it is fun to look at the maps and B-factors near the ligand. The B-factors go way up and the negative difference density disappears, as does most of the 2mFo-DFc density. It’s obviously not news to most people on the CCP4BB, properly refining B-factors is really important! Cheers, Robbie From: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] On Behalf Of Bärbel Blaum Sent: Friday, July 19, 2019 17:24 To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Hi Rhys, the reported B-factors for the “ligands” are all way below the reported B-factors for the protein chains, with the worst of the three complexes reporting unitless numbers 23.2 and 64.8, respectively, just to highlight *one* striking feature of the data collection and refinement table. So even with the limited info normally available to reviewers this table should have raised a red flag. After the re-refinement suggested by others, i.e. your own proper assessment of the crystallographic data, if you do not find noteworthy density you may want to contact the article’s editor with your results. If you work in this field, i.e. really care about this paper scientifically and you are not afraid to confront the authors you could suggest writing a comment/direct response article but in my opinion that would only make sense if you can be sure beforehand that it will be linked visibly to the actual paper, else it will be a waste of time. And don’t forget that just because one or some of the authors did a bad job at the crystallographic end other findings
Re: [ccp4bb] AW: [EXTERNAL] [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Dear Kay and colleagues, It is possible to extend the inspection of data associated to a PDB entry beyond Table 1 by means of the PDBpeep server available at http://staraniso.globalphasing.org/cgi-bin/PDBpeep.cgi Doing so for these three entries shows that while 6MO2 has a rather normal 3D distribution of ) and , both 6MO0 and 6MO1 show a strange pattern for these two quantities, where the iso-surfaces are pinched towards the origin along the c* axis - perhaps, indeed, an indication that the space group is wrong and that the point group is 3 rather than 32 (?). It might be something else, but there is a definite anomaly in the 3D distributions of these two statistics. With best wishes, Gerard. -- On Fri, Jul 19, 2019 at 06:07:33PM +0100, Kay Diederichs wrote: > Hi Tristan et al, > > before people invest their time into re-refining a model against the > deposited data, I'd suggest that these data are dubious - see table S3 in > https://pubs.acs.org/doi/suppl/10.1021/jacs.9b02505/suppl_file/ja9b02505_si_001.pdf > ! The data processing statistics look "surprising" to me: "Rsym or Rmerge" > values are unexpectedly high, and the high-resolution cutoffs for 2 of the 3 > structures are too low. Maybe the space group is wrong? With such data, > ligand density cannot expected to be meaningful, I'd say. > In my eyes, the authors should be asked for the raw data images, and these > should be reprocessed before re-refinement. > > best, > Kay > > On Fri, 19 Jul 2019 17:05:21 +0100, Tristan Croll wrote: > > >I couldn't resist doing a little playing with 6mo0 in ISOLDE. The > >sequence past Gly1151 is wrong - after jumping a small gap (density > >clearly indicates a single missing residue here), it continues on from > >1165. Corrected the following sequence to GVVTRS, deleted the ligands, > >did some basic energy minimisation and cleaning up of obvious local > >problems, reset the B-factors and refined in Phenix.refine. Rwork/Rfree > >28/32, MolProbity score 2.14 (vs. 3.6 originally). No sign of anything > >like the ligand in the difference density (just a small blob in the > >active site). Some other oddities make me suspect a number of point > >mutations from the assigned sequence (not register errors - the > >surrounds are quite clear). I'm particularly intrigued by residues 48-71 > >- this stretch appears to link seamlessly head-to-tail with its own > >symmetry copy, forming a seemingly-continuous chain snaking through the > >crystal. Can't make head nor tail of that. A decidedly odd structure... > >but I agree with you that the ligand doesn't seem supported by the > >density. > > > >Best regards, > > > >Tristan > > > >On 2019-07-19 14:46, herman.schreu...@sanofi.com wrote: > >> Hi Rhys, > >> > >> There is definitively some density present for a ligand, but the > >> active site region looks completely misfitted, and the ligand density > >> may also belong to unfitted protein residues. One first needs to get > >> the protein chain right, and should then look if there would still be > >> density available to fit the ligand. > >> > >> Best, > >> > >> Herman > >> > >> VON: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] IM AUFTRAG VON > >> Rhys Grinter > >> GESENDET: Freitag, 19. Juli 2019 15:22 > >> AN: CCP4BB@JISCMAIL.AC.UK > >> BETREFF: [EXTERNAL] [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, > >> 6MO2 > >> > >> EXTERNAL : Real sender is owner-ccp...@jiscmail.ac.uk > >> > >> Hi All, > >> > >> I was chatting with a colleague during a recent synchrotron visit and > >> they'd recently come across some ligand/drug bound structures > >> associated with a paper recently published in a high impact factor > >> journal. > >> > >> They had pulled the associated SFs from the PDB and found that the > >> electron density associated with these ligands didn't match that > >> reported in the paper and certainly wasn't sufficient to model the > >> alleged ligand. > >> > >> I also pulled the structure factors and after refinement in the > >> presence/absence of the alleged ligand I also feel that the density > >> present does not warrant modelling of the ligand. > >> > >> I was hoping that the community might be able to give me an outside > >> opinion on these datasets (PDB IDs: 6MO0, 6MO1, 6MO2) and if the > >> problem associated with the data is verified, provide some ad
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
I like cases like this as cautionary tails for the crystallography class. In this case, the sequence is mismatch in several places, ex. residues 1165-1170 should be 1153-1158 which flows nicely after you add a residue to 1151. There are other places where the mainchain has slid and is out of registry. Other places where the sequence is mismatched or doesn't match the plasmid, ex. 1108. The ligand not withstanding, the structure could be cleaned up in a few hours. I'll add these to the validation class in the fall. Best- Todd From: CCP4 bulletin board on behalf of Robbie Joosten Sent: Friday, July 19, 2019 1:24 PM To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Even though PDB-REDO cannot salvage this model without extensive rebuilding which is what Tristan showed, it is fun to look at the maps and B-factors near the ligand. The B-factors go way up and the negative difference density disappears, as does most of the 2mFo-DFc density. It’s obviously not news to most people on the CCP4BB, properly refining B-factors is really important! Cheers, Robbie From: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] On Behalf Of Bärbel Blaum Sent: Friday, July 19, 2019 17:24 To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Hi Rhys, the reported B-factors for the “ligands” are all way below the reported B-factors for the protein chains, with the worst of the three complexes reporting unitless numbers 23.2 and 64.8, respectively, just to highlight *one* striking feature of the data collection and refinement table. So even with the limited info normally available to reviewers this table should have raised a red flag. After the re-refinement suggested by others, i.e. your own proper assessment of the crystallographic data, if you do not find noteworthy density you may want to contact the article’s editor with your results. If you work in this field, i.e. really care about this paper scientifically and you are not afraid to confront the authors you could suggest writing a comment/direct response article but in my opinion that would only make sense if you can be sure beforehand that it will be linked visibly to the actual paper, else it will be a waste of time. And don’t forget that just because one or some of the authors did a bad job at the crystallographic end other findings of the paper might still be solid – in collaborations often one author is unable to critically evaluate another author’s contribution and this would not be the first case were good synthetic or biological work is presented along with a bad crystal structure. By the way and a bit ironically this protein may have suffered bad crystallography/scientific practice before - I think it was one of the fake Krishna Murthy structures, right? The associated (now retracted) article I mean is here https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub Kind regards, Bärbel --- Bärbel Blaum, PhD Inthera Bioscience AG Einsiedlerstrasse 34 CH-8820 Waedenswil Switzerland E-Mail: baerbel.bl...@intherabio.com Phone: +41 43 477 94 72-- Von: CCP4 bulletin board im Auftrag von "Manfred S. Weiss" Antworten an: "Manfred S. Weiss" Datum: Freitag, 19. Juli 2019 um 16:03 An: Betreff: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Hi Rhys, all three structures are at modest resolution and they don't seem to be properly refined. At least they are all below average. I wonder how this paper made it past the referees. I haven't checked the paper, but there are ways and means how to deal with weakly bound ligands in the best possible way. One aspect is to improve the phases as much as possible without having the ligand present. This was obviously NOT done. Another way is to use the PANDDA approach, which relies on having many data sets available. I suppose that this was also not done. The best way to check is to delete the ligand and so some extensive refinement in order to remove the phase bias introduced by the ligand. Only then you can reliably assess whether something is there or not. Cheers, Manfred Am 19.07.2019 um 15:21 schrieb Rhys Grinter: Hi All, I was chatting with a colleague during a recent synchrotron visit and they'd recently come across some ligand/drug bound structures associated with a paper recently published in a high impact factor journal. They had pulled the associated SFs from the PDB and found that the electron density associated with these ligands didn't match that reported in the paper and certainly wasn't sufficient to model the alleged ligand. I also pulled the structure factors and after refinement in the presence/absence of the alleged ligand I also feel that the density present does not warrant modelling of the ligand. I was hoping that the c
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
I don’t think anyone mentioned contacting the authors first—doesn’t this seem like the first thing one should do? Jacob + Jacob Pearson Keller Research Scientist / Looger Lab HHMI Janelia Research Campus 19700 Helix Dr, Ashburn, VA 20147 Desk: (571)209-4000 x3159 Cell: (301)592-7004 + The content of this email is confidential and intended for the recipient specified in message only. It is strictly forbidden to share any part of this message with any third party, without a written consent of the sender. If you received this message by mistake, please reply to this message and follow with its deletion, so that we can ensure such a mistake does not occur in the future. From: CCP4 bulletin board On Behalf Of Roger Rowlett Sent: Friday, July 19, 2019 4:48 PM To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 If it is as it appears, it is disappointing to see this in JACS. I would expect better. Unfortunately, reviewers don't always get a lot of information to judge quality of structures (which has been discussed extensively on prior occasions on this board), so some trust is involved that what data is presented by the authors is representative. I see the authors duly presented a 2Fo-Fc map and a Fo-Fc omit map in one of the figures, but it appears from a personal examination that the Fo-Fc omit map presented may not resemble what one gets when personally examining the deposited data, and some protein chain refinement problems are immediately evident. That is a bit concerning. High-pressure research, coupled with (perhaps) a lack of understanding, experience, and supervision is a potentially dangerous ethical stew. Correcting erroneous published data is challenging. I share the frustration of others when attempting to challenge scientifically questionable published results. My trip down the rabbit hole (not related directly to crystallography) nearly identical to that related previously here. Spend your own time writing a full paper rebuttal (ugh) or move on and concentrate resources on your own work. _ Roger S. Rowlett Gordon & Dorothy Kline Professor, Emeritus Department of Chemistry Colgate University email: rrowl...@colgate.edu<mailto:rrowl...@colgate.edu> On 7/19/2019 2:24 PM, Robbie Joosten wrote: Even though PDB-REDO cannot salvage this model without extensive rebuilding which is what Tristan showed, it is fun to look at the maps and B-factors near the ligand. The B-factors go way up and the negative difference density disappears, as does most of the 2mFo-DFc density. It’s obviously not news to most people on the CCP4BB, properly refining B-factors is really important! Cheers, Robbie From: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] On Behalf Of Bärbel Blaum Sent: Friday, July 19, 2019 17:24 To: CCP4BB@JISCMAIL.AC.UK<mailto:CCP4BB@JISCMAIL.AC.UK> Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Hi Rhys, the reported B-factors for the “ligands” are all way below the reported B-factors for the protein chains, with the worst of the three complexes reporting unitless numbers 23.2 and 64.8, respectively, just to highlight *one* striking feature of the data collection and refinement table. So even with the limited info normally available to reviewers this table should have raised a red flag. After the re-refinement suggested by others, i.e. your own proper assessment of the crystallographic data, if you do not find noteworthy density you may want to contact the article’s editor with your results. If you work in this field, i.e. really care about this paper scientifically and you are not afraid to confront the authors you could suggest writing a comment/direct response article but in my opinion that would only make sense if you can be sure beforehand that it will be linked visibly to the actual paper, else it will be a waste of time. And don’t forget that just because one or some of the authors did a bad job at the crystallographic end other findings of the paper might still be solid – in collaborations often one author is unable to critically evaluate another author’s contribution and this would not be the first case were good synthetic or biological work is presented along with a bad crystal structure. By the way and a bit ironically this protein may have suffered bad crystallography/scientific practice before - I think it was one of the fake Krishna Murthy structures, right? The associated (now retracted) article I mean is here https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub<https://urldefense.proofpoint.com/v2/url?u=https-3A__www.sciencedirect.com_science_article_pii_S002228360093924X-3Fvia-253Dihub&d=DwMDaQ&c=LU6cRtx0xgB8s29tIz9Olw&r=eLCg9eJ4R
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
If it is as it appears, it is disappointing to see this in JACS. I would expect better. Unfortunately, reviewers don't always get a lot of information to judge quality of structures (which has been discussed extensively on prior occasions on this board), so some trust is involved that what data is presented by the authors is representative. I see the authors duly presented a 2Fo-Fc map and a Fo-Fc omit map in one of the figures, but it appears from a personal examination that the Fo-Fc omit map presented may not resemble what one gets when personally examining the deposited data, and some protein chain refinement problems are immediately evident. That is a bit concerning. High-pressure research, coupled with (perhaps) a lack of understanding, experience, and supervision is a potentially dangerous ethical stew. Correcting erroneous published data is challenging. I share the frustration of others when attempting to challenge scientifically questionable published results. My trip down the rabbit hole (not related directly to crystallography) nearly identical to that related previously here. Spend your own time writing a full paper rebuttal (ugh) or move on and concentrate resources on your own work. _ Roger S. Rowlett Gordon & Dorothy Kline Professor, Emeritus Department of Chemistry Colgate University email: rrowl...@colgate.edu On 7/19/2019 2:24 PM, Robbie Joosten wrote: Even though PDB-REDO cannot salvage this model without extensive rebuilding which is what Tristan showed, it is fun to look at the maps and B-factors near the ligand. The B-factors go way up and the negative difference density disappears, as does most of the 2mFo-DFc density. It’s obviously not news to most people on the CCP4BB, properly refining B-factors is really important! Cheers, Robbie *From:*CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] *On Behalf Of *Bärbel Blaum *Sent:* Friday, July 19, 2019 17:24 *To:* CCP4BB@JISCMAIL.AC.UK *Subject:* Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Hi Rhys, the reported B-factors for the “ligands” are all way below the reported B-factors for the protein chains, with the worst of the three complexes reporting unitless numbers 23.2 and 64.8, respectively, just to highlight *one* striking feature of the data collection and refinement table. So even with the limited info normally available to reviewers this table should have raised a red flag. After the re-refinement suggested by others, i.e. your own proper assessment of the crystallographic data, if you do not find noteworthy density you may want to contact the article’s editor with your results. If you work in this field, i.e. really care about this paper scientifically and you are not afraid to confront the authors you could suggest writing a comment/direct response article but in my opinion that would only make sense if you can be sure beforehand that it will be linked visibly to the actual paper, else it will be a waste of time. And don’t forget that just because one or some of the authors did a bad job at the crystallographic end other findings of the paper might still be solid – in collaborations often one author is unable to critically evaluate another author’s contribution and this would not be the first case were good synthetic or biological work is presented along with a bad crystal structure. By the way and a bit ironically this protein may have suffered bad crystallography/scientific practice before - I think it was one of the fake Krishna Murthy structures, right? The associated (now retracted) article I mean is here https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub Kind regards, Bärbel --- Bärbel Blaum, PhD Inthera Bioscience AG Einsiedlerstrasse 34 CH-8820 Waedenswil Switzerland E-Mail: baerbel.bl...@intherabio.com Phone: +41 43 477 94 72-- *Von: *CCP4 bulletin board im Auftrag von "Manfred S. Weiss" *Antworten an: *"Manfred S. Weiss" *Datum: *Freitag, 19. Juli 2019 um 16:03 *An: * *Betreff: *Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Hi Rhys, all three structures are at modest resolution and they don't seem to be properly refined. At least they are all below average. I wonder how this paper made it past the referees. I haven't checked the paper, but there are ways and means how to deal with weakly bound ligands in the best possible way. One aspect is to improve the phases as much as possible without having the ligand present. This was obviously NOT done. Another way is to use the PANDDA approach, which relies on having many data sets available. I suppose that this was also not done. The best way to check is to delete the ligand and so some extensive refinement in order to remove the phase bias introduced by the ligand. Only then you can reliably assess whether something is there or not. Cheers, Manfred Am 1
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Even though PDB-REDO cannot salvage this model without extensive rebuilding which is what Tristan showed, it is fun to look at the maps and B-factors near the ligand. The B-factors go way up and the negative difference density disappears, as does most of the 2mFo-DFc density. It’s obviously not news to most people on the CCP4BB, properly refining B-factors is really important! Cheers, Robbie From: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] On Behalf Of Bärbel Blaum Sent: Friday, July 19, 2019 17:24 To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Hi Rhys, the reported B-factors for the “ligands” are all way below the reported B-factors for the protein chains, with the worst of the three complexes reporting unitless numbers 23.2 and 64.8, respectively, just to highlight *one* striking feature of the data collection and refinement table. So even with the limited info normally available to reviewers this table should have raised a red flag. After the re-refinement suggested by others, i.e. your own proper assessment of the crystallographic data, if you do not find noteworthy density you may want to contact the article’s editor with your results. If you work in this field, i.e. really care about this paper scientifically and you are not afraid to confront the authors you could suggest writing a comment/direct response article but in my opinion that would only make sense if you can be sure beforehand that it will be linked visibly to the actual paper, else it will be a waste of time. And don’t forget that just because one or some of the authors did a bad job at the crystallographic end other findings of the paper might still be solid – in collaborations often one author is unable to critically evaluate another author’s contribution and this would not be the first case were good synthetic or biological work is presented along with a bad crystal structure. By the way and a bit ironically this protein may have suffered bad crystallography/scientific practice before - I think it was one of the fake Krishna Murthy structures, right? The associated (now retracted) article I mean is here https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub Kind regards, Bärbel --- Bärbel Blaum, PhD Inthera Bioscience AG Einsiedlerstrasse 34 CH-8820 Waedenswil Switzerland E-Mail: baerbel.bl...@intherabio.com Phone: +41 43 477 94 72-- Von: CCP4 bulletin board im Auftrag von "Manfred S. Weiss" Antworten an: "Manfred S. Weiss" Datum: Freitag, 19. Juli 2019 um 16:03 An: Betreff: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Hi Rhys, all three structures are at modest resolution and they don't seem to be properly refined. At least they are all below average. I wonder how this paper made it past the referees. I haven't checked the paper, but there are ways and means how to deal with weakly bound ligands in the best possible way. One aspect is to improve the phases as much as possible without having the ligand present. This was obviously NOT done. Another way is to use the PANDDA approach, which relies on having many data sets available. I suppose that this was also not done. The best way to check is to delete the ligand and so some extensive refinement in order to remove the phase bias introduced by the ligand. Only then you can reliably assess whether something is there or not. Cheers, Manfred Am 19.07.2019 um 15:21 schrieb Rhys Grinter: Hi All, I was chatting with a colleague during a recent synchrotron visit and they'd recently come across some ligand/drug bound structures associated with a paper recently published in a high impact factor journal. They had pulled the associated SFs from the PDB and found that the electron density associated with these ligands didn't match that reported in the paper and certainly wasn't sufficient to model the alleged ligand. I also pulled the structure factors and after refinement in the presence/absence of the alleged ligand I also feel that the density present does not warrant modelling of the ligand. I was hoping that the community might be able to give me an outside opinion on these datasets (PDB IDs: 6MO0, 6MO1, 6MO2) and if the problem associated with the data is verified, provide some advice on how to proceed. This isn't the first occasion I've seen ligand bound structures with questionable density deposited in association with papers in well respected journals. Despite improvements to validation I feel that this problem is widespread. Best Regards, Rhys -- Dr Rhys Grinter NHMRC Postdoctoral Researcher Monash University +61 (0)3 9902 9213 +61 (0)403 896 767 To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1 -- Dr. Manfred S. Weiss Macromolecular Crystallography Helmho
[ccp4bb] Fwd: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Begin forwarded message: From: "charles w carter, jr" mailto:cwcar...@ad.unc.edu>> Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Date: July 19, 2019 at 1:32:53 PM EDT To: Patrick Loll mailto:pjl...@gmail.com>> Hi Pat, I, too have fallen into this rabbit hole while Mark Rould and I were writing a Methods in Enzymology chapter on Fobs - Fobs difference Fourier maps as a vastly better tool for documenting ligands. Mark had documented a similar offense, and I had previously also encountered one. Both cases occurred since structure factor deposition was frequent and/or required, so it was relatively trivial to do the definitive experiments. The density of outright fraudulent results of this type in high profile journals long ago persuaded me that unless someone pointed me to an especially salient article, that those journals did not merit my attention, and I’ve several times refused to review manuscripts for them, for much the same reason. I was recently asked to teach a module on scientific fraud in a newly required course on ethics for NIH Programmatic grants. It was both amusing and rewarding, as the students readily understood many of these issues. “Scholarly Publishing” is something best left to professional societies in my (admittedly extreme) opinion. Charlie On Jul 19, 2019, at 1:17 PM, Patrick Loll mailto:pjl...@gmail.com>> wrote: The idea of contacting the editor (and/or author) is an excellent one, and indeed the correct thing to do scientifically. However, I’m disillusioned: I’ve been down this path before with a high-profile vanity journal, and while the editors paid lip service to the notion that the record should be corrected, in reality they led me on for the better part of a year, and got me to write up detailed analyses of why the ligand positioning was not justified, before eventually saying “no, we don’t see any need to publish a correction.” I speculate that the journal prefers not avoid corrections, for fear that too many corrections will make the journal a less desirable destination. On 19 Jul 2019, at 11:23 AM, Bärbel Blaum mailto:baerbel.bl...@intherabio.com>> wrote: Hi Rhys, the reported B-factors for the “ligands” are all way below the reported B-factors for the protein chains, with the worst of the three complexes reporting unitless numbers 23.2 and 64.8, respectively, just to highlight *one* striking feature of the data collection and refinement table. So even with the limited info normally available to reviewers this table should have raised a red flag. After the re-refinement suggested by others, i.e. your own proper assessment of the crystallographic data, if you do not find noteworthy density you may want to contact the article’s editor with your results. If you work in this field, i.e. really care about this paper scientifically and you are not afraid to confront the authors you could suggest writing a comment/direct response article but in my opinion that would only make sense if you can be sure beforehand that it will be linked visibly to the actual paper, else it will be a waste of time. And don’t forget that just because one or some of the authors did a bad job at the crystallographic end other findings of the paper might still be solid – in collaborations often one author is unable to critically evaluate another author’s contribution and this would not be the first case were good synthetic or biological work is presented along with a bad crystal structure. By the way and a bit ironically this protein may have suffered bad crystallography/scientific practice before - I think it was one of the fake Krishna Murthy structures, right? The associated (now retracted) article I mean is here https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub Kind regards, Bärbel --- Bärbel Blaum, PhD Inthera Bioscience AG Einsiedlerstrasse 34 CH-8820 Waedenswil Switzerland E-Mail: baerbel.bl...@intherabio.com Phone: +41 43 477 94 72-- Von: CCP4 bulletin board im Auftrag von "Manfred S. Weiss" Antworten an: "Manfred S. Weiss" Datum: Freitag, 19. Juli 2019 um 16:03 An: Betreff: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Hi Rhys, all three structures are at modest resolution and they don't seem to be properly refined. At least they are all below average. I wonder how this paper made it past the referees. I haven't checked the paper, but there are ways and means how to deal with weakly bound ligands in the best possible way. One aspect is to improve the phases as much as possible without having the ligand present. This was obviously NOT done. Another way is to use the PANDDA approach, which relies on having many data sets available. I suppose that this was also not done. The best way to check is to delete the ligand and so some extensive refinement in order to remove the phase bias introduced by the ligand.
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Would it be possible to add a public annotations section to the PDB, to allow us to potentially flag/warn whoever downloads that particular structure, there could be something wrong with it, such as wrong space group, no/poor density fitting for ligand. Something similar to PubPeer maybe? Daniel A. Bonsor PhD Institute of Human Virology, University of Maryland at Baltimore 725 W. Lombard Street N571 Baltimore MD 21201 From: CCP4 bulletin board on behalf of Patrick Loll Sent: Friday, July 19, 2019 1:17 PM To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 The idea of contacting the editor (and/or author) is an excellent one, and indeed the correct thing to do scientifically. However, I’m disillusioned: I’ve been down this path before with a high-profile vanity journal, and while the editors paid lip service to the notion that the record should be corrected, in reality they led me on for the better part of a year, and got me to write up detailed analyses of why the ligand positioning was not justified, before eventually saying “no, we don’t see any need to publish a correction.” I speculate that the journal prefers not avoid corrections, for fear that too many corrections will make the journal a less desirable destination. > On 19 Jul 2019, at 11:23 AM, Bärbel Blaum > wrote: > > Hi Rhys, > the reported B-factors for the “ligands” are all way below the reported > B-factors for the protein chains, with the worst of the three complexes > reporting unitless numbers 23.2 and 64.8, respectively, just to highlight > *one* striking feature of the data collection and refinement table. So even > with the limited info normally available to reviewers this table should have > raised a red flag. After the re-refinement suggested by others, i.e. your own > proper assessment of the crystallographic data, if you do not find noteworthy > density you may want to contact the article’s editor with your results. If > you work in this field, i.e. really care about this paper scientifically and > you are not afraid to confront the authors you could suggest writing a > comment/direct response article but in my opinion that would only make sense > if you can be sure beforehand that it will be linked visibly to the actual > paper, else it will be a waste of time. And don’t forget that just because > one or some of the authors did a bad job at the crystallographic end other > findings of the paper might still be solid – in collaborations often one > author is unable to critically evaluate another author’s contribution and > this would not be the first case were good synthetic or biological work is > presented along with a bad crystal structure. > By the way and a bit ironically this protein may have suffered bad > crystallography/scientific practice before - I think it was one of the fake > Krishna Murthy structures, right? The associated (now retracted) article I > mean is here > https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub [https://ars.els-cdn.com/content/image/S00222836.gif]<https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub> RETRACTED: Crystal structure of dengue virus NS3 protease in complex with a bowman-birk inhibitor: implications for flaviviral polyprotein processing and drug design - ScienceDirect - ScienceDirect.com | Science, health and medical journals, full text articles and books.<https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub> www.sciencedirect.com COMMUNIC Crystal Structure of Dengue Complex with a Bowman-Bir ro L. 1Center for Macromolecular C f A 8 U T MCLM 244, Birmingham AL 35294-0005, USA 2Department of Biochemistry and Molecular Biology, Kansas University Medical Center 3901 Rainbow Boulevard Kansas City, KS 66160- 7421, USA Dengue viruses are members of the Flaviviridae and cause dengue fever Dengue fever and dengue hemorrhagic ... > Kind regards, Bärbel > --- > Bärbel Blaum, PhD > Inthera Bioscience AG > Einsiedlerstrasse 34 > CH-8820 Waedenswil > Switzerland > E-Mail: baerbel.bl...@intherabio.com > Phone: +41 43 477 94 72-- > > > > Von: CCP4 bulletin board im Auftrag von "Manfred S. > Weiss" > Antworten an: "Manfred S. Weiss" > Datum: Freitag, 19. Juli 2019 um 16:03 > An: > Betreff: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 > > Hi Rhys, > > all three structures are at modest resolution and they don't seem to > be properly refined. At least they are all below average. I wonder > how this paper made it past the referees. > > I haven't checked the paper, but there are ways and means how to > deal with weakly bound ligands in the best possible way. One aspect > is to imp
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
The idea of contacting the editor (and/or author) is an excellent one, and indeed the correct thing to do scientifically. However, I’m disillusioned: I’ve been down this path before with a high-profile vanity journal, and while the editors paid lip service to the notion that the record should be corrected, in reality they led me on for the better part of a year, and got me to write up detailed analyses of why the ligand positioning was not justified, before eventually saying “no, we don’t see any need to publish a correction.” I speculate that the journal prefers not avoid corrections, for fear that too many corrections will make the journal a less desirable destination. > On 19 Jul 2019, at 11:23 AM, Bärbel Blaum > wrote: > > Hi Rhys, > the reported B-factors for the “ligands” are all way below the reported > B-factors for the protein chains, with the worst of the three complexes > reporting unitless numbers 23.2 and 64.8, respectively, just to highlight > *one* striking feature of the data collection and refinement table. So even > with the limited info normally available to reviewers this table should have > raised a red flag. After the re-refinement suggested by others, i.e. your own > proper assessment of the crystallographic data, if you do not find noteworthy > density you may want to contact the article’s editor with your results. If > you work in this field, i.e. really care about this paper scientifically and > you are not afraid to confront the authors you could suggest writing a > comment/direct response article but in my opinion that would only make sense > if you can be sure beforehand that it will be linked visibly to the actual > paper, else it will be a waste of time. And don’t forget that just because > one or some of the authors did a bad job at the crystallographic end other > findings of the paper might still be solid – in collaborations often one > author is unable to critically evaluate another author’s contribution and > this would not be the first case were good synthetic or biological work is > presented along with a bad crystal structure. > By the way and a bit ironically this protein may have suffered bad > crystallography/scientific practice before - I think it was one of the fake > Krishna Murthy structures, right? The associated (now retracted) article I > mean is here > https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub > Kind regards, Bärbel > --- > Bärbel Blaum, PhD > Inthera Bioscience AG > Einsiedlerstrasse 34 > CH-8820 Waedenswil > Switzerland > E-Mail: baerbel.bl...@intherabio.com > Phone: +41 43 477 94 72-- > > > > Von: CCP4 bulletin board im Auftrag von "Manfred S. > Weiss" > Antworten an: "Manfred S. Weiss" > Datum: Freitag, 19. Juli 2019 um 16:03 > An: > Betreff: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 > > Hi Rhys, > > all three structures are at modest resolution and they don't seem to > be properly refined. At least they are all below average. I wonder > how this paper made it past the referees. > > I haven't checked the paper, but there are ways and means how to > deal with weakly bound ligands in the best possible way. One aspect > is to improve the phases as much as possible without having the ligand > present. This was obviously NOT done. Another way is to use the > PANDDA approach, which relies on having many data sets available. > I suppose that this was also not done. > > The best way to check is to delete the ligand and so some extensive > refinement in order to remove the phase bias introduced by the > ligand. Only then you can reliably assess whether something is there > or not. > > Cheers, Manfred > > Am 19.07.2019 um 15:21 schrieb Rhys Grinter: >> Hi All, >> >> I was chatting with a colleague during a recent synchrotron visit and they'd >> recently come across some ligand/drug bound structures associated with a >> paper recently published in a high impact factor journal. >> >> They had pulled the associated SFs from the PDB and found that the electron >> density associated with these ligands didn't match that reported in the >> paper and certainly wasn't sufficient to model the alleged ligand. >> >> I also pulled the structure factors and after refinement in the >> presence/absence of the alleged ligand I also feel that the density present >> does not warrant modelling of the ligand. >> >> I was hoping that the community might be able to give me an outside opinion >> on these datasets (PDB IDs: 6MO0, 6MO1, 6MO2) and if the problem associated >> with the data is verified, provide some advi
Re: [ccp4bb] AW: [EXTERNAL] [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Hi Tristan et al, before people invest their time into re-refining a model against the deposited data, I'd suggest that these data are dubious - see table S3 in https://pubs.acs.org/doi/suppl/10.1021/jacs.9b02505/suppl_file/ja9b02505_si_001.pdf ! The data processing statistics look "surprising" to me: "Rsym or Rmerge" values are unexpectedly high, and the high-resolution cutoffs for 2 of the 3 structures are too low. Maybe the space group is wrong? With such data, ligand density cannot expected to be meaningful, I'd say. In my eyes, the authors should be asked for the raw data images, and these should be reprocessed before re-refinement. best, Kay On Fri, 19 Jul 2019 17:05:21 +0100, Tristan Croll wrote: >I couldn't resist doing a little playing with 6mo0 in ISOLDE. The >sequence past Gly1151 is wrong - after jumping a small gap (density >clearly indicates a single missing residue here), it continues on from >1165. Corrected the following sequence to GVVTRS, deleted the ligands, >did some basic energy minimisation and cleaning up of obvious local >problems, reset the B-factors and refined in Phenix.refine. Rwork/Rfree >28/32, MolProbity score 2.14 (vs. 3.6 originally). No sign of anything >like the ligand in the difference density (just a small blob in the >active site). Some other oddities make me suspect a number of point >mutations from the assigned sequence (not register errors - the >surrounds are quite clear). I'm particularly intrigued by residues 48-71 >- this stretch appears to link seamlessly head-to-tail with its own >symmetry copy, forming a seemingly-continuous chain snaking through the >crystal. Can't make head nor tail of that. A decidedly odd structure... >but I agree with you that the ligand doesn't seem supported by the >density. > >Best regards, > >Tristan > >On 2019-07-19 14:46, herman.schreu...@sanofi.com wrote: >> Hi Rhys, >> >> There is definitively some density present for a ligand, but the >> active site region looks completely misfitted, and the ligand density >> may also belong to unfitted protein residues. One first needs to get >> the protein chain right, and should then look if there would still be >> density available to fit the ligand. >> >> Best, >> >> Herman >> >> VON: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] IM AUFTRAG VON >> Rhys Grinter >> GESENDET: Freitag, 19. Juli 2019 15:22 >> AN: CCP4BB@JISCMAIL.AC.UK >> BETREFF: [EXTERNAL] [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, >> 6MO2 >> >> EXTERNAL : Real sender is owner-ccp...@jiscmail.ac.uk >> >> Hi All, >> >> I was chatting with a colleague during a recent synchrotron visit and >> they'd recently come across some ligand/drug bound structures >> associated with a paper recently published in a high impact factor >> journal. >> >> They had pulled the associated SFs from the PDB and found that the >> electron density associated with these ligands didn't match that >> reported in the paper and certainly wasn't sufficient to model the >> alleged ligand. >> >> I also pulled the structure factors and after refinement in the >> presence/absence of the alleged ligand I also feel that the density >> present does not warrant modelling of the ligand. >> >> I was hoping that the community might be able to give me an outside >> opinion on these datasets (PDB IDs: 6MO0, 6MO1, 6MO2) and if the >> problem associated with the data is verified, provide some advice on >> how to proceed. >> >> This isn't the first occasion I've seen ligand bound structures with >> questionable density deposited in association with papers in well >> respected journals. Despite improvements to validation I feel that >> this problem is widespread. >> >> Best Regards, >> >> Rhys >> >> -- >> >> Dr Rhys Grinter >> >> NHMRC Postdoctoral Researcher >> >> Monash University >> >> +61 (0)3 9902 9213 >> >> +61 (0)403 896 767 >> >> - >> >> To unsubscribe from the CCP4BB list, click the following link: >> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1 [1] >> >> - >> >> To unsubscribe from the CCP4BB list, click the following link: >> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1 [2] >> >> Links: >> -- >> [1] >> https://urldefense.proofpoint.com/v2/url?u=https-3A__www.jiscmail.ac.uk_cgi-2Dbin_webadmin-3FSUBED1-3DCCP4BB-26A-3D1&d=DwMFaQ&c=Dbf9zoswcQ-C
Re: [ccp4bb] AW: [EXTERNAL] [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
I couldn't resist doing a little playing with 6mo0 in ISOLDE. The sequence past Gly1151 is wrong - after jumping a small gap (density clearly indicates a single missing residue here), it continues on from 1165. Corrected the following sequence to GVVTRS, deleted the ligands, did some basic energy minimisation and cleaning up of obvious local problems, reset the B-factors and refined in Phenix.refine. Rwork/Rfree 28/32, MolProbity score 2.14 (vs. 3.6 originally). No sign of anything like the ligand in the difference density (just a small blob in the active site). Some other oddities make me suspect a number of point mutations from the assigned sequence (not register errors - the surrounds are quite clear). I'm particularly intrigued by residues 48-71 - this stretch appears to link seamlessly head-to-tail with its own symmetry copy, forming a seemingly-continuous chain snaking through the crystal. Can't make head nor tail of that. A decidedly odd structure... but I agree with you that the ligand doesn't seem supported by the density. Best regards, Tristan On 2019-07-19 14:46, herman.schreu...@sanofi.com wrote: Hi Rhys, There is definitively some density present for a ligand, but the active site region looks completely misfitted, and the ligand density may also belong to unfitted protein residues. One first needs to get the protein chain right, and should then look if there would still be density available to fit the ligand. Best, Herman VON: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] IM AUFTRAG VON Rhys Grinter GESENDET: Freitag, 19. Juli 2019 15:22 AN: CCP4BB@JISCMAIL.AC.UK BETREFF: [EXTERNAL] [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 EXTERNAL : Real sender is owner-ccp...@jiscmail.ac.uk Hi All, I was chatting with a colleague during a recent synchrotron visit and they'd recently come across some ligand/drug bound structures associated with a paper recently published in a high impact factor journal. They had pulled the associated SFs from the PDB and found that the electron density associated with these ligands didn't match that reported in the paper and certainly wasn't sufficient to model the alleged ligand. I also pulled the structure factors and after refinement in the presence/absence of the alleged ligand I also feel that the density present does not warrant modelling of the ligand. I was hoping that the community might be able to give me an outside opinion on these datasets (PDB IDs: 6MO0, 6MO1, 6MO2) and if the problem associated with the data is verified, provide some advice on how to proceed. This isn't the first occasion I've seen ligand bound structures with questionable density deposited in association with papers in well respected journals. Despite improvements to validation I feel that this problem is widespread. Best Regards, Rhys -- Dr Rhys Grinter NHMRC Postdoctoral Researcher Monash University +61 (0)3 9902 9213 +61 (0)403 896 767 - To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1 [1] - To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1 [2] Links: -- [1] https://urldefense.proofpoint.com/v2/url?u=https-3A__www.jiscmail.ac.uk_cgi-2Dbin_webadmin-3FSUBED1-3DCCP4BB-26A-3D1&d=DwMFaQ&c=Dbf9zoswcQ-CRvvI7VX5j3HvibIuT3ZiarcKl5qtMPo&r=HK-CY_tL8CLLA93vdywyu3qI70R4H8oHzZyRHMQu1AQ&m=KR65gleAznAqJKiHPp9vvVZeSsv8LHvr5PDyckgEaWw&s=NmR2g-LBQtuhDFsFRJPffbzIm3cQQWMhK7ozfrse7KY&e= [2] https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1 To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Hi Rhys, the reported B-factors for the “ligands” are all way below the reported B-factors for the protein chains, with the worst of the three complexes reporting unitless numbers 23.2 and 64.8, respectively, just to highlight *one* striking feature of the data collection and refinement table. So even with the limited info normally available to reviewers this table should have raised a red flag. After the re-refinement suggested by others, i.e. your own proper assessment of the crystallographic data, if you do not find noteworthy density you may want to contact the article’s editor with your results. If you work in this field, i.e. really care about this paper scientifically and you are not afraid to confront the authors you could suggest writing a comment/direct response article but in my opinion that would only make sense if you can be sure beforehand that it will be linked visibly to the actual paper, else it will be a waste of time. And don’t forget that just because one or some of the authors did a bad job at the crystallographic end other findings of the paper might still be solid – in collaborations often one author is unable to critically evaluate another author’s contribution and this would not be the first case were good synthetic or biological work is presented along with a bad crystal structure. By the way and a bit ironically this protein may have suffered bad crystallography/scientific practice before - I think it was one of the fake Krishna Murthy structures, right? The associated (now retracted) article I mean is here https://www.sciencedirect.com/science/article/pii/S002228360093924X?via%3Dihub Kind regards, Bärbel --- Bärbel Blaum, PhD Inthera Bioscience AG Einsiedlerstrasse 34 CH-8820 Waedenswil Switzerland E-Mail: baerbel.bl...@intherabio.com Phone: +41 43 477 94 72-- Von: CCP4 bulletin board im Auftrag von "Manfred S. Weiss" Antworten an: "Manfred S. Weiss" Datum: Freitag, 19. Juli 2019 um 16:03 An: Betreff: Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 Hi Rhys, all three structures are at modest resolution and they don't seem to be properly refined. At least they are all below average. I wonder how this paper made it past the referees. I haven't checked the paper, but there are ways and means how to deal with weakly bound ligands in the best possible way. One aspect is to improve the phases as much as possible without having the ligand present. This was obviously NOT done. Another way is to use the PANDDA approach, which relies on having many data sets available. I suppose that this was also not done. The best way to check is to delete the ligand and so some extensive refinement in order to remove the phase bias introduced by the ligand. Only then you can reliably assess whether something is there or not. Cheers, Manfred Am 19.07.2019 um 15:21 schrieb Rhys Grinter: Hi All, I was chatting with a colleague during a recent synchrotron visit and they'd recently come across some ligand/drug bound structures associated with a paper recently published in a high impact factor journal. They had pulled the associated SFs from the PDB and found that the electron density associated with these ligands didn't match that reported in the paper and certainly wasn't sufficient to model the alleged ligand. I also pulled the structure factors and after refinement in the presence/absence of the alleged ligand I also feel that the density present does not warrant modelling of the ligand. I was hoping that the community might be able to give me an outside opinion on these datasets (PDB IDs: 6MO0, 6MO1, 6MO2) and if the problem associated with the data is verified, provide some advice on how to proceed. This isn't the first occasion I've seen ligand bound structures with questionable density deposited in association with papers in well respected journals. Despite improvements to validation I feel that this problem is widespread. Best Regards, Rhys -- Dr Rhys Grinter NHMRC Postdoctoral Researcher Monash University +61 (0)3 9902 9213 +61 (0)403 896 767 To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1 -- Dr. Manfred S. Weiss Macromolecular Crystallography Helmholtz-Zentrum Berlin Albert-Einstein-Str. 15 D-12489 Berlin Germany Helmholtz-Zentrum Berlin für Materialien und Energie GmbH Mitglied der Hermann von Helmholtz-Gemeinschaft Deutscher Forschungszentren e.V. Aufsichtsrat: Vorsitzender Dr. Volkmar Dietz, stv. Vorsitzende Dr. Jutta Koch-Unterseher Geschäftsführung: Prof. Dr. Bernd Rech (Sprecher), Prof. Dr. Jan Lüning, Thomas Frederking Sitz Berlin, AG Charlottenburg, 89 HRB 5583 Postadresse: Hahn-Meitner-Platz 1 D-14109 Berlin To unsubscribe from the CCP4BB list, click the following link: ht
Re: [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Hi Rhys, all three structures are at modest resolution and they don't seem to be properly refined. At least they are all below average. I wonder how this paper made it past the referees. I haven't checked the paper, but there are ways and means how to deal with weakly bound ligands in the best possible way. One aspect is to improve the phases as much as possible without having the ligand present. This was obviously NOT done. Another way is to use the PANDDA approach, which relies on having many data sets available. I suppose that this was also not done. The best way to check is to delete the ligand and so some extensive refinement in order to remove the phase bias introduced by the ligand. Only then you can reliably assess whether something is there or not. Cheers, Manfred Am 19.07.2019 um 15:21 schrieb Rhys Grinter: Hi All, I was chatting with a colleague during a recent synchrotron visit and they'd recently come across some ligand/drug bound structures associated with a paper recently published in a high impact factor journal. They had pulled the associated SFs from the PDB and found that the electron density associated with these ligands didn't match that reported in the paper and certainly wasn't sufficient to model the alleged ligand. I also pulled the structure factors and after refinement in the presence/absence of the alleged ligand I also feel that the density present does not warrant modelling of the ligand. I was hoping that the community might be able to give me an outside opinion on these datasets (PDB IDs: 6MO0, 6MO1, 6MO2) and if the problem associated with the data is verified, provide some advice on how to proceed. This isn't the first occasion I've seen ligand bound structures with questionable density deposited in association with papers in well respected journals. Despite improvements to validation I feel that this problem is widespread. Best Regards, Rhys -- Dr Rhys Grinter NHMRC Postdoctoral Researcher Monash University +61 (0)3 9902 9213 +61 (0)403 896 767 To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1 -- Dr. Manfred S. Weiss Macromolecular Crystallography Helmholtz-Zentrum Berlin Albert-Einstein-Str. 15 D-12489 Berlin Germany Helmholtz-Zentrum Berlin für Materialien und Energie GmbH Mitglied der Hermann von Helmholtz-Gemeinschaft Deutscher Forschungszentren e.V. Aufsichtsrat: Vorsitzender Dr. Volkmar Dietz, stv. Vorsitzende Dr. Jutta Koch-Unterseher Geschäftsführung: Prof. Dr. Bernd Rech (Sprecher), Prof. Dr. Jan Lüning, Thomas Frederking Sitz Berlin, AG Charlottenburg, 89 HRB 5583 Postadresse: Hahn-Meitner-Platz 1 D-14109 Berlin To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1
Re: [ccp4bb] [EXTERNAL] [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
The validation reports are pretty bad, so it seems to be a case where the referees and editor have not checked them. And, like Herman wrote, the models do not appear to agree well enough with the maps - in all three cases. Mark J van Raaij Dpto de Estructura de Macromoleculas Centro Nacional de Biotecnologia - CSIC calle Darwin 3 E-28049 Madrid, Spain tel. (+34) 91 585 4616 > On 19 Jul 2019, at 15:46, herman.schreu...@sanofi.com wrote: > > Hi Rhys, > > There is definitively some density present for a ligand, but the active site > region looks completely misfitted, and the ligand density may also belong to > unfitted protein residues. One first needs to get the protein chain right, > and should then look if there would still be density available to fit the > ligand. > > Best, > Herman > > Von: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK > <mailto:CCP4BB@JISCMAIL.AC.UK>] Im Auftrag von Rhys Grinter > Gesendet: Freitag, 19. Juli 2019 15:22 > An: CCP4BB@JISCMAIL.AC.UK <mailto:CCP4BB@JISCMAIL.AC.UK> > Betreff: [EXTERNAL] [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 > > EXTERNAL : Real sender is owner-ccp...@jiscmail.ac.uk > <mailto:owner-ccp...@jiscmail.ac.uk> > > > Hi All, > > I was chatting with a colleague during a recent synchrotron visit and they'd > recently come across some ligand/drug bound structures associated with a > paper recently published in a high impact factor journal. > > They had pulled the associated SFs from the PDB and found that the electron > density associated with these ligands didn't match that reported in the paper > and certainly wasn't sufficient to model the alleged ligand. > > I also pulled the structure factors and after refinement in the > presence/absence of the alleged ligand I also feel that the density present > does not warrant modelling of the ligand. > > I was hoping that the community might be able to give me an outside opinion > on these datasets (PDB IDs: 6MO0, 6MO1, 6MO2) and if the problem associated > with the data is verified, provide some advice on how to proceed. > > This isn't the first occasion I've seen ligand bound structures with > questionable density deposited in association with papers in well respected > journals. Despite improvements to validation I feel that this problem is > widespread. > > Best Regards, > > Rhys > > -- > Dr Rhys Grinter > NHMRC Postdoctoral Researcher > Monash University > +61 (0)3 9902 9213 > +61 (0)403 896 767 > > To unsubscribe from the CCP4BB list, click the following link: > https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1 > <https://urldefense.proofpoint.com/v2/url?u=https-3A__www.jiscmail.ac.uk_cgi-2Dbin_webadmin-3FSUBED1-3DCCP4BB-26A-3D1&d=DwMFaQ&c=Dbf9zoswcQ-CRvvI7VX5j3HvibIuT3ZiarcKl5qtMPo&r=HK-CY_tL8CLLA93vdywyu3qI70R4H8oHzZyRHMQu1AQ&m=KR65gleAznAqJKiHPp9vvVZeSsv8LHvr5PDyckgEaWw&s=NmR2g-LBQtuhDFsFRJPffbzIm3cQQWMhK7ozfrse7KY&e=> > To unsubscribe from the CCP4BB list, click the following link: > https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1 > <https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1> To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1
[ccp4bb] AW: [EXTERNAL] [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Hi Rhys, There is definitively some density present for a ligand, but the active site region looks completely misfitted, and the ligand density may also belong to unfitted protein residues. One first needs to get the protein chain right, and should then look if there would still be density available to fit the ligand. Best, Herman Von: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] Im Auftrag von Rhys Grinter Gesendet: Freitag, 19. Juli 2019 15:22 An: CCP4BB@JISCMAIL.AC.UK Betreff: [EXTERNAL] [ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2 EXTERNAL : Real sender is owner-ccp...@jiscmail.ac.uk Hi All, I was chatting with a colleague during a recent synchrotron visit and they'd recently come across some ligand/drug bound structures associated with a paper recently published in a high impact factor journal. They had pulled the associated SFs from the PDB and found that the electron density associated with these ligands didn't match that reported in the paper and certainly wasn't sufficient to model the alleged ligand. I also pulled the structure factors and after refinement in the presence/absence of the alleged ligand I also feel that the density present does not warrant modelling of the ligand. I was hoping that the community might be able to give me an outside opinion on these datasets (PDB IDs: 6MO0, 6MO1, 6MO2) and if the problem associated with the data is verified, provide some advice on how to proceed. This isn't the first occasion I've seen ligand bound structures with questionable density deposited in association with papers in well respected journals. Despite improvements to validation I feel that this problem is widespread. Best Regards, Rhys -- Dr Rhys Grinter NHMRC Postdoctoral Researcher Monash University +61 (0)3 9902 9213 +61 (0)403 896 767 To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1<https://urldefense.proofpoint.com/v2/url?u=https-3A__www.jiscmail.ac.uk_cgi-2Dbin_webadmin-3FSUBED1-3DCCP4BB-26A-3D1&d=DwMFaQ&c=Dbf9zoswcQ-CRvvI7VX5j3HvibIuT3ZiarcKl5qtMPo&r=HK-CY_tL8CLLA93vdywyu3qI70R4H8oHzZyRHMQu1AQ&m=KR65gleAznAqJKiHPp9vvVZeSsv8LHvr5PDyckgEaWw&s=NmR2g-LBQtuhDFsFRJPffbzIm3cQQWMhK7ozfrse7KY&e=> To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1
[ccp4bb] Questionable Ligand Density: 6MO0, 6MO1, 6MO2
Hi All, I was chatting with a colleague during a recent synchrotron visit and they'd recently come across some ligand/drug bound structures associated with a paper recently published in a high impact factor journal. They had pulled the associated SFs from the PDB and found that the electron density associated with these ligands didn't match that reported in the paper and certainly wasn't sufficient to model the alleged ligand. I also pulled the structure factors and after refinement in the presence/absence of the alleged ligand I also feel that the density present does not warrant modelling of the ligand. I was hoping that the community might be able to give me an outside opinion on these datasets (PDB IDs: 6MO0, 6MO1, 6MO2) and if the problem associated with the data is verified, provide some advice on how to proceed. This isn't the first occasion I've seen ligand bound structures with questionable density deposited in association with papers in well respected journals. Despite improvements to validation I feel that this problem is widespread. Best Regards, Rhys -- Dr Rhys Grinter NHMRC Postdoctoral Researcher Monash University +61 (0)3 9902 9213 +61 (0)403 896 767 To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1
