Re: [ccp4bb] RES: [ccp4bb] Se-SAD phasing

[Holton, James M]

It would not surprise me at all if pseudotranslation (aka tNCS) were holding 
you back.  It is like kryptonite to all phasing techniques, MR, MAD, SAD, MIR, 
you name it.  I recently did an analysis of old, never-solved data sets from my 
beamline.  These are MAD/SAD data that don't match anything in the PDB, despite 
being 10 years old or more.  At least 45% of them suffer from 
pseudotranslation.  Another 21% are twinned.

In contrast, if you search the PDB for the word "twin" only 0.5% of all entries 
contain that string.  Might explain why methods development on these two fronts 
has been slow.

I don't suppose you'd be up for crowdsourcing?  As in making your raw data 
publicly available?

-James Holton
MAD Scientist

On 7/18/2019 11:34 AM, Mario Tyago Murakami wrote:
Curiously, in the native data we observed a strong peak of 28%. However, the 
same was not seen in the Semet crystals. There is isomorphism between these 
crystals. Please see below the unit cell parameters. In the previous email I 
sent attached pointless, index.lp from xds and shelxc logs.

Native -> UNIT_CELL_CONSTANTS=   118.082   151.969   166.266  90.000  90.000  
90.000 (tNCS 28% F.C. = 0.5 0.135 0.50)
Semet  -> UNIT_CELL_CONSTANTS=   118.730   151.831   167.070  90.000  90.000  
90.000 (tNCS 8%  F.C. = 0.5 0.140 0.50)


De: Eleanor Dodson 
Enviada em: quinta-feira, 18 de julho de 2019 12:34
Para: Mario Tyago Murakami 

Cc: CCP4BB@JISCMAIL.AC.UK
Assunto: Re: [ccp4bb] Se-SAD phasing

Good advice above..
Someting extra - I would look at information about how the molecules might be 
arranged.
Is there any non-crystallographic translation? ( One with a coordinate 0f say 
z= 0.5 could mean the spacegroup
could be P21212 or P212121 )

Sometimes there is a sub-cell and you can search for fewer sites in that cell..
Eleanor



On Thu, 18 Jul 2019 at 16:28, Mario Tyago Murakami 
mailto:mario.murak...@lnbr.cnpem.br>> wrote:
Dear Andy, Clemens and Michal

The solution for orthrombic SG is very clear. P1 indexing results in a very 
similar Unit cell parameters without appreciable deviations in angles 
<0.1degrees.

It seems more likely the problem pointed by Andy of a lot of sites. I will try 
all the suggestions and hope to come back soon with good news.

Thanks
Mario

Em 18 de jul de 2019 11:44, THOMPSON Andrew 
mailto:andrew.thomp...@synchrotron-soleil.fr>>
 escreveu:

Dear Mario

The data seem very good but you are looking for an awful lot of sites. Keep 
trying with SHELX, but there are some keywords that you might use which may 
help (see extract from SHELX doc). This worked for me for a couple of 
structures with > 100 sites. You should also try with and without Patterson 
seeding.  You may need many thousand trials…. I would also start with the data 
cutoff somewhere between 3.5 - 4 A (to get rid of anisotropy) . You could also 
try manually modifying the Emin value to use just your strongest signal. Don’t 
expect the solution to come out easily - you may have to try different shelx 
runs for days before getting a single solution.

At the same time, you really do need to be absolutely certain of the space 
group ……any tests you can do

Good luck

Andy





For large selenomethionine substructures (which behave more like equal atom ab 
initiostructure solution of small molecules) it may be worth increasing the 
number of Pattersonpeaks used for the Patterson seeding (e.g. PATT 200; the 
default is 100) and adding theinstructions WEED 0.3 (random omit maps) and SKIP 
0.5 (uranium atom removal). Thelatter two are the defaults when PLOP is present 
but are switched off by default if PLOP isabsent. When PATS is used, WEED 
produces a much smaller additional improvement in thehit ratio than when PATS 
is absent. For small substructures (<10 sites), WEED and SKIP cando more harm 
than good by eliminating too many correct sites at once.



De : CCP4 bulletin board 
[mailto:CCP4BB@JISCMAIL.AC.UK] De la part de 
Mario Tyago Murakami
Envoyé : jeudi 18 juillet 2019 16:19
À : CCP4BB@JISCMAIL.AC.UK
Objet : [ccp4bb] Se-SAD phasing



Dear all,

I am trying to solve the phases of the following SeMet data, but so far 
unsuccessfully. Suggestions are very welcome. Please see below some details 
about the case.



The statistics below is from a merged data from  different kappas of the same 
crystal to increase redundancy. We used the fixed energy 12675 eV since the 
fluorescence detector was not working at the used beamline to get best energy 
values for this crystal.

Xtriage did not indicate any crystallographic pathology, except moderate 
anisotropy.

The unit cells parameters are 118.72   151.82   167.05  90.000  90.000  90.000 
(P212121) containing from 8 to 12 molecules in the asymmetric unit. The protein 
has ~28.5 kDa and 10 Met residues, excluding those from the N- and C-termini, 
proba

Re: [ccp4bb] RES: [ccp4bb] Se-SAD phasing

[Tim Gruene]

Dear Mario,

you probably want to run more than NTRY 1000. With today's computers, NTRY 
5000 or NTRY 1 should not take very long.

You high resolution cut-off in shelxd, 2.7A, is the default, as shelxc adds 
0.5A to the high resolution limit of the full data set. In your case this 
seems reasonable, as CC(1/2) drops below 30% at about 2.75A. 

You are looking for Se incorporated into SeMet, i.e. your scatterers cannot 
have a special position. Set the second number of MIND to e.g. 3.5 (>0!) to 
reject the possibility that heavy atoms have special positions.

The main reason. why you may not have found a solution could the the FIND 8 
command. This tells shelxd to look for 8 Se-atoms. You describe yourself that 
you are expecting 80-120 Se-atom. So set with FIND 100

I strongly recommend using the GUI hkl2map by Fabio Dall'Antonio, http://
webapps.embl-hamburg.de/hkl2map/ It guides your very well through the steps of 
shelx c/d/e 

As for the space group, your indexing in XDS looks reasonable, I don't think 
that XDS accidentally doubled one of the unit cell axes.

However, from pointless.log, the decision for the 2(1) b-axis is based on 26 
reflections. I would normally trust pointless, and expect that you get as 
solution by properly setting the parameters in shelxd and trying autobuilding 
in shelxe (your resolution limit for shelxe is probably 1.9-2.0; 2.2 was quite 
a cautious resolution cut-off).

However, you may want to re-run the CORRECT steps with
SPACE_GROUP_NUMBER=16
UNIT_CELL_CONSTANTS= 118.836 167.150 151.908 90 90 90   
REIDX= -1 0 0 0  0 0 1 0  0 1 0 0

and run shelx c/d/e with spacegroup P2(1)2(1)2

(sorry, I am not sure wether conventionally a Sure, it follows attached. I am also attaching the pointless log and
> IDXREF.LP from xds.
 
> -Mensagem original-
> De: Tim Grüne  
> Enviada em: quinta-feira, 18 de julho de 2019 13:06
> Para: Mario Tyago Murakami 
> Cc: CCP4BB@jiscmail.ac.uk
> Assunto: Re: [ccp4bb] Se-SAD phasing
> 
> Dear Mario,
> 
> you may want to post the shelxc log file and the shelxd input file to get a
> bit more quantitative advice with shelxd.  For the spacegroup, it is worth
> reading the pointless log file carefully about how many expected absences
> you have per screw axis, although pointless is quite careful and does not
> suggest a screw axes when the respective axis was not recorded.
 
> 
> 
>  From your CCanom, you could cut your resolution to 2.8A for shelxd (your
> native data probably goes to 2.0'ish A, by the way).
 
> Best regards,
> Tim
> 
> Am 18.07.2019 16:19, schrieb Mario Tyago Murakami:
> 
> > Dear all,
> > 
> > I am trying to solve the phases of the following SeMet data, but so 
> > far unsuccessfully. Suggestions are very welcome. Please see below 
> > some details about the case.
> > 
> > The statistics below is from a merged data from  different kappas of 
> > the same crystal to increase redundancy. We used the fixed energy
> > 12675 eV since the fluorescence detector was not working at the used 
> > beamline to get best energy values for this crystal.
> > 
> > Xtriage did not indicate any crystallographic pathology, except 
> > moderate anisotropy.
> > 
> > The unit cells parameters are 118.72   151.82   167.05  90.000  90.000
> > 
> >  90.000 (P212121) containing from 8 to 12 molecules in the asymmetric 
> > 
> > unit. The protein has ~28.5 kDa and 10 Met residues, excluding those 
> > from the N- and C-termini, probably with low occupancy. Thus, 
> > something 80 to 120 scatterers are expected.
> > 
> > Phenix_anomalous_signal indicates a probability of 99% to solve it and 
> > the anomalous signal is theoretically in a very good range.
> > 
> > I have tried SHELXD with different resolutions and number of sites. I 
> > have also used Hyss. But all attempts failed.
> > 
> > Thanks in advance
> > 
> > Mario
> > 
> > Aviso Legal: Esta mensagem e seus anexos podem conter informações 
> > confidenciais e/ou de uso restrito. Observe atentamente seu conteúdo e 
> > considere eventual consulta ao remetente antes de copiá-la, divulgá-la 
> > ou distribuí-la. Se você recebeu esta mensagem por engano, por favor 
> > avise o remetente e apague-a imediatamente.
> > 
> > Disclaimer: This email and its attachments may contain confidential 
> > and/or privileged information. Observe its content carefully and 
> > consider possible querying to the sender before copying, disclosing or 
> > distributing it. If you have received this email by mistake, please 
> > notify the sender and delete it immediately.
> > 
> > -
> > 
> > To unsubscribe from the CCP4BB list, click the following link:
> > https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1
> 
> 
> --
> --
> Tim Gruene
> Head of the Centre for X-ray Structure Analysis Faculty of Chemistry
> University of Vienna
 
> Phone: +43-1-4277-70202
> 
> GPG Key ID = A46BEE1A

-- 
--
Tim Gruene
Head of the Centre for X-ray Structure Analysis
Faculty of Chemistry
University of V

[ccp4bb] RES: [ccp4bb] Se-SAD phasing

[Mario Tyago Murakami]

Curiously, in the native data we observed a strong peak of 28%. However, the 
same was not seen in the Semet crystals. There is isomorphism between these 
crystals. Please see below the unit cell parameters. In the previous email I 
sent attached pointless, index.lp from xds and shelxc logs.

Native -> UNIT_CELL_CONSTANTS=   118.082   151.969   166.266  90.000  90.000  
90.000 (tNCS 28% F.C. = 0.5 0.135 0.50)
Semet  -> UNIT_CELL_CONSTANTS=   118.730   151.831   167.070  90.000  90.000  
90.000 (tNCS 8%  F.C. = 0.5 0.140 0.50)


De: Eleanor Dodson 
Enviada em: quinta-feira, 18 de julho de 2019 12:34
Para: Mario Tyago Murakami 
Cc: CCP4BB@JISCMAIL.AC.UK
Assunto: Re: [ccp4bb] Se-SAD phasing

Good advice above..
Someting extra - I would look at information about how the molecules might be 
arranged.
Is there any non-crystallographic translation? ( One with a coordinate 0f say 
z= 0.5 could mean the spacegroup
could be P21212 or P212121 )

Sometimes there is a sub-cell and you can search for fewer sites in that cell..
Eleanor



On Thu, 18 Jul 2019 at 16:28, Mario Tyago Murakami 
mailto:mario.murak...@lnbr.cnpem.br>> wrote:
Dear Andy, Clemens and Michal

The solution for orthrombic SG is very clear. P1 indexing results in a very 
similar Unit cell parameters without appreciable deviations in angles 
<0.1degrees.

It seems more likely the problem pointed by Andy of a lot of sites. I will try 
all the suggestions and hope to come back soon with good news.

Thanks
Mario

Em 18 de jul de 2019 11:44, THOMPSON Andrew 
mailto:andrew.thomp...@synchrotron-soleil.fr>>
 escreveu:

Dear Mario

The data seem very good but you are looking for an awful lot of sites. Keep 
trying with SHELX, but there are some keywords that you might use which may 
help (see extract from SHELX doc). This worked for me for a couple of 
structures with > 100 sites. You should also try with and without Patterson 
seeding.  You may need many thousand trials…. I would also start with the data 
cutoff somewhere between 3.5 - 4 A (to get rid of anisotropy) . You could also 
try manually modifying the Emin value to use just your strongest signal. Don’t 
expect the solution to come out easily - you may have to try different shelx 
runs for days before getting a single solution.

At the same time, you really do need to be absolutely certain of the space 
group ……any tests you can do

Good luck

Andy





For large selenomethionine substructures (which behave more like equal atom ab 
initiostructure solution of small molecules) it may be worth increasing the 
number of Pattersonpeaks used for the Patterson seeding (e.g. PATT 200; the 
default is 100) and adding theinstructions WEED 0.3 (random omit maps) and SKIP 
0.5 (uranium atom removal). Thelatter two are the defaults when PLOP is present 
but are switched off by default if PLOP isabsent. When PATS is used, WEED 
produces a much smaller additional improvement in thehit ratio than when PATS 
is absent. For small substructures (<10 sites), WEED and SKIP cando more harm 
than good by eliminating too many correct sites at once.



De : CCP4 bulletin board 
[mailto:CCP4BB@JISCMAIL.AC.UK] De la part de 
Mario Tyago Murakami
Envoyé : jeudi 18 juillet 2019 16:19
À : CCP4BB@JISCMAIL.AC.UK
Objet : [ccp4bb] Se-SAD phasing



Dear all,

I am trying to solve the phases of the following SeMet data, but so far 
unsuccessfully. Suggestions are very welcome. Please see below some details 
about the case.



The statistics below is from a merged data from  different kappas of the same 
crystal to increase redundancy. We used the fixed energy 12675 eV since the 
fluorescence detector was not working at the used beamline to get best energy 
values for this crystal.

Xtriage did not indicate any crystallographic pathology, except moderate 
anisotropy.

The unit cells parameters are 118.72   151.82   167.05  90.000  90.000  90.000 
(P212121) containing from 8 to 12 molecules in the asymmetric unit. The protein 
has ~28.5 kDa and 10 Met residues, excluding those from the N- and C-termini, 
probably with low occupancy. Thus, something 80 to 120 scatterers are expected.

Phenix_anomalous_signal indicates a probability of 99% to solve it and the 
anomalous signal is theoretically in a very good range.

I have tried SHELXD with different resolutions and number of sites. I have also 
used Hyss. But all attempts failed.



Thanks in advance

Mario





Aviso Legal: Esta mensagem e seus anexos podem conter informações confidenciais 
e/ou de uso restrito. Observe atentamente seu conteúdo e considere eventual 
consulta ao remetente antes de copiá-la, divulgá-la ou distribuí-la. Se você 
recebeu esta mensagem por engano, por favor avise o remetente e apague-a 
imediatamente.

Disclaimer: This email and its attachments may contain confidential and/or 
privileged information. Observe its content carefully and consider possible 
querying to the sender before copying, disclosin

[ccp4bb] RES: [ccp4bb] Se-SAD phasing

[Mario Tyago Murakami]

Sure, it follows attached. I am also attaching the pointless log and IDXREF.LP 
from xds.

-Mensagem original-
De: Tim Grüne  
Enviada em: quinta-feira, 18 de julho de 2019 13:06
Para: Mario Tyago Murakami 
Cc: CCP4BB@jiscmail.ac.uk
Assunto: Re: [ccp4bb] Se-SAD phasing

Dear Mario,

you may want to post the shelxc log file and the shelxd input file to get a bit 
more quantitative advice with shelxd.  For the spacegroup, it is worth reading 
the pointless log file carefully about how many expected absences you have per 
screw axis, although pointless is quite careful and does not suggest a screw 
axes when the respective axis was not recorded.



 From your CCanom, you could cut your resolution to 2.8A for shelxd (your 
native data probably goes to 2.0'ish A, by the way).

Best regards,
Tim

Am 18.07.2019 16:19, schrieb Mario Tyago Murakami:
> Dear all,
> 
> I am trying to solve the phases of the following SeMet data, but so 
> far unsuccessfully. Suggestions are very welcome. Please see below 
> some details about the case.
> 
> The statistics below is from a merged data from  different kappas of 
> the same crystal to increase redundancy. We used the fixed energy
> 12675 eV since the fluorescence detector was not working at the used 
> beamline to get best energy values for this crystal.
> 
> Xtriage did not indicate any crystallographic pathology, except 
> moderate anisotropy.
> 
> The unit cells parameters are 118.72   151.82   167.05  90.000  90.000
>  90.000 (P212121) containing from 8 to 12 molecules in the asymmetric 
> unit. The protein has ~28.5 kDa and 10 Met residues, excluding those 
> from the N- and C-termini, probably with low occupancy. Thus, 
> something 80 to 120 scatterers are expected.
> 
> Phenix_anomalous_signal indicates a probability of 99% to solve it and 
> the anomalous signal is theoretically in a very good range.
> 
> I have tried SHELXD with different resolutions and number of sites. I 
> have also used Hyss. But all attempts failed.
> 
> Thanks in advance
> 
> Mario
> 
> Aviso Legal: Esta mensagem e seus anexos podem conter informações 
> confidenciais e/ou de uso restrito. Observe atentamente seu conteúdo e 
> considere eventual consulta ao remetente antes de copiá-la, divulgá-la 
> ou distribuí-la. Se você recebeu esta mensagem por engano, por favor 
> avise o remetente e apague-a imediatamente.
> 
> Disclaimer: This email and its attachments may contain confidential 
> and/or privileged information. Observe its content carefully and 
> consider possible querying to the sender before copying, disclosing or 
> distributing it. If you have received this email by mistake, please 
> notify the sender and delete it immediately.
> 
> -
> 
> To unsubscribe from the CCP4BB list, click the following link:
> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1

--
--
Tim Gruene
Head of the Centre for X-ray Structure Analysis Faculty of Chemistry University 
of Vienna

Phone: +43-1-4277-70202

GPG Key ID = A46BEE1A



To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1


sad_shelxc.log
Description: sad_shelxc.log


pointless.log
Description: pointless.log


IDXREF.LP
Description: IDXREF.LP