Hi Shya, I did wavelength scan, got a good signal for Se and used appropriate wavelengths for data collection and also used experimental f' and f'' values for phasing. I think the reasons SAD or MAD for SeMet data is not working are (i) low resolution: 3.7 for SeMet (Anomalous data is up to 4.8A) (ii) I should have mentioned this earlier: 3 out of 6 Se are very close to N-terminus, possible they are disordered. Unit cell is also some what big..100, 120 and 320A; F222 space group. AK -----Original Message----- From: Shya Biswas To: PHENIX user mailing list Sent: Thu, Sep 6, 2012 7:29 am Subject: Re: [phenixbb] Autosol: MIRAS Hi AK, Did you do a wavelength scan when you collected the SE dataset you need to put the values of f' and f'' from your wavelength scan in order to locate the heavy atom sites, 6 methionine should be enough to phase your molecule. Shya On Wed, Sep 5, 2012 at 9:25 PM, wrote:

Hi all,

I am trying to solve a structure through experimental phasing using AUTOSOL. I have a couple of heavy atom derivative datasets (Hg, La, Eu, Cd) and also a SeMet data. Unfortunately all the datasets are of low resolution (3.7-4.2A) and there are possibly 4-8 molecules in the asu. MIR, SAD and MAD alone did not give any convincing solution.

However, MIRAS, with a combination of few heavy atom datasets and the anomalous data from SeMet crystals, gave a somewhat satisfactory map. But the heavy atom site picked by AUTOSOL list only one of the heavy atoms i.e. Lanthanum. In another set of run, the solution of which was not convincing, the heavy atom substructure had only Hg. There are 6 Met out of 200 residues in one molecule and mass spec results show that Se incorporation is 100%.

Now, my doubt is that why does the heavy atom substructure contain only La and how can I get the substructure involving Se from this solution (or the datasets used)? Se location is going to help me a lot for finding a starting point to assign side chains.

Any suggestion would be greatly appreciated.

Thanks AK

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

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

Satisfactory map means: I am expecting a coiled coil or helix bundle type of assembly and I could see some densities appropriate for helices. There are proper solvent channels and continuous stretches of densities. More to add about data on this: the data is anisotropic and the longer 'c' axis and alignment of helical density along c axis support this. This also makes me think that perhaps map is sensible. Few rounds of model building, refinement and DM has been successful to assign around 10 polyala helices and their distribution looks sensible from packing point of view. Now the problem is to assign the side chain. I was hoping to make use of Se locations for this. R and Rfree are still random, in the range of 50%. I am not sure, but it could be perhaps because most of the scattering material is still unassigned. -----Original Message----- From: Francis E Reyes To: PHENIX user mailing list Cc: phenixbb Sent: Thu, Sep 6, 2012 8:07 am Subject: Re: [phenixbb] Autosol: MIRAS This one is solvable, but with extreme difficulty. I recently completed a structure solution with experimental phases starting at 5.0 A using phase information from multiple derivatives. How would you describe a somewhat satisfactory map? F On Sep 5, 2012, at 7:08 PM, [email protected] wrote: Hi Shya, I did wavelength scan, got a good signal for Se and used appropriate wavelengths for data collection and also used experimental f' and f'' values for phasing. I think the reasons SAD or MAD for SeMet data is not working are (i) low resolution: 3.7 for SeMet (Anomalous data is up to 4.8A) (ii) I should have mentioned this earlier: 3 out of 6 Se are very close to N-terminus, possible they are disordered. Unit cell is also some what big..100, 120 and 320A; F222 space group. AK -----Original Message----- From: Shya Biswas To: PHENIX user mailing list Sent: Thu, Sep 6, 2012 7:29 am Subject: Re: [phenixbb] Autosol: MIRAS Hi AK, Did you do a wavelength scan when you collected the SE dataset you need to put the values of f' and f'' from your wavelength scan in order to locate the heavy atom sites, 6 methionine should be enough to phase your molecule. Shya On Wed, Sep 5, 2012 at 9:25 PM, wrote:

Hi all,

I am trying to solve a structure through experimental phasing using AUTOSOL. I have a couple of heavy atom derivative datasets (Hg, La, Eu, Cd) and also a SeMet data. Unfortunately all the datasets are of low resolution (3.7-4.2A) and there are possibly 4-8 molecules in the asu. MIR, SAD and MAD alone did not give any convincing solution.

However, MIRAS, with a combination of few heavy atom datasets and the anomalous data from SeMet crystals, gave a somewhat satisfactory map. But the heavy atom site picked by AUTOSOL list only one of the heavy atoms i.e. Lanthanum. In another set of run, the solution of which was not convincing, the heavy atom substructure had only Hg. There are 6 Met out of 200 residues in one molecule and mass spec results show that Se incorporation is 100%.

Now, my doubt is that why does the heavy atom substructure contain only La and how can I get the substructure involving Se from this solution (or the datasets used)? Se location is going to help me a lot for finding a starting point to assign side chains.

Any suggestion would be greatly appreciated.

Thanks AK

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

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

@Gabor: Thanks Gabor! I'll try MR-SAD. Hope it works. @Andre: I have tried Shelx and Phenix. Se sites from both of them were not the same. I tried giving Shelxd output to Phenix also. That also did not help. I have not tried Autosharp yet. I'll see that too.. However, what now I am primarily looking at is the way to get Se-substructure from the MIRAS run. Is there a way to get that? autosol gives only one heavy atom in the output! Since I have a map which looks promising, so before going to the output of another run I want to explore all the possibilities from this. Se locations in this map should help. AK -----Original Message----- From: Gabor Bunkoczi To: PHENIX user mailing list Sent: Thu, Sep 6, 2012 5:52 am Subject: Re: [phenixbb] Autosol: MIRAS Dear AK, if you just need the Se positions, you can run Phaser MR-SAD (available from the GUI) using your partial model as a starting point, together with the Se-Met data. Dependending on the model quality (and the model does not have to be very good), it may be able to locate the Se-sites. BW, Gabor On Sep 6 2012, [email protected] wrote:

Satisfactory map means: I am expecting a coiled coil or helix bundle type of assembly and I could see some densities appropriate for helices. There are proper solvent channels and continuous stretches of densities.

More to add about data on this: the data is anisotropic and the longer 'c' axis and alignment of helical density along c axis support this. This also makes me think that perhaps map is sensible.

Few rounds of model building, refinement and DM has been successful to assign around 10 polyala helices and their distribution looks sensible from packing point of view. Now the problem is to assign the side chain. I was hoping to make use of Se locations for this. R and Rfree are still random, in the range of 50%. I am not sure, but it could be perhaps because most of the scattering material is still unassigned.

-----Original Message----- From: Francis E Reyes To: PHENIX user mailing list Cc: phenixbb Sent: Thu, Sep 6, 2012 8:07 am Subject: Re: [phenixbb] Autosol: MIRAS

This one is solvable, but with extreme difficulty. I recently completed a structure solution with experimental phases starting at 5.0 A using phase information from multiple derivatives.

How would you describe a somewhat satisfactory map?

F

On Sep 5, 2012, at 7:08 PM, [email protected] wrote:

Hi Shya,

I did wavelength scan, got a good signal for Se and used appropriate wavelengths for data collection and also used experimental f' and f'' values for phasing. I think the reasons SAD or MAD for SeMet data is not working are (i) low resolution: 3.7 for SeMet (Anomalous data is up to 4.8A) (ii) I should have mentioned this earlier: 3 out of 6 Se are very close to N-terminus, possible they are disordered. Unit cell is also some what big..100, 120 and 320A; F222 space group.

AK

-----Original Message----- From: Shya Biswas To: PHENIX user mailing list Sent: Thu, Sep 6, 2012 7:29 am Subject: Re: [phenixbb] Autosol: MIRAS

Hi AK, Did you do a wavelength scan when you collected the SE dataset you need to put the values of f' and f'' from your wavelength scan in order to locate the heavy atom sites, 6 methionine should be enough to phase your molecule. Shya

On Wed, Sep 5, 2012 at 9:25 PM, wrote:

...

Hi all,

I am trying to solve a structure through experimental phasing using AUTOSOL. I have a couple of heavy atom derivative datasets (Hg, La, Eu, Cd) and also a SeMet data. Unfortunately all the datasets are of low resolution (3.7-4.2A) and there are possibly 4-8 molecules in the asu. MIR, SAD and MAD alone did not give any convincing solution.

However, MIRAS, with a combination of few heavy atom datasets and the anomalous data from SeMet crystals, gave a somewhat satisfactory map. But the heavy atom site picked by AUTOSOL list only one of the heavy atoms i.e. Lanthanum. In another set of run, the solution of which was not convincing, the heavy atom substructure had only Hg. There are 6 Met out of 200 residues in one molecule and mass spec results show that Se incorporation is 100%.

Now, my doubt is that why does the heavy atom substructure contain only La and how can I get the substructure involving Se from this solution (or the datasets used)? Se location is going to help me a lot for finding a starting point to assign side chains.

Any suggestion would be greatly appreciated.

Thanks AK

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

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

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

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

-- ################################################## Dr Gabor Bunkoczi Cambridge Institute for Medical Research Wellcome Trust/MRC Building Addenbrooke's Hospital Hills Road Cambridge CB2 0XY ################################################## _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

Yup, I have tried anisotropy truncation... -----Original Message----- From: Shya Biswas To: PHENIX user mailing list Sent: Thu, Sep 6, 2012 8:32 am Subject: Re: [phenixbb] Autosol: MIRAS Hi AK, Generally Se incorporation is not good at the N-terminus but since you have six you should be able to locate the other three even at 4.8A. Did you try using the anisotropy server, this might help you improve the map, here's the reference: http://www.pnas.org/content/103/21/8060.full Shya On Wed, Sep 5, 2012 at 10:08 PM, wrote:

Hi Shya,

I did wavelength scan, got a good signal for Se and used appropriate wavelengths for data collection and also used experimental f' and f'' values for phasing. I think the reasons SAD or MAD for SeMet data is not working are (i) low resolution: 3.7 for SeMet (Anomalous data is up to 4.8A) (ii) I should have mentioned this earlier: 3 out of 6 Se are very close to N-terminus, possible they are disordered. Unit cell is also some what big..100, 120 and 320A; F222 space group.

AK

-----Original Message----- From: Shya Biswas To: PHENIX user mailing list Sent: Thu, Sep 6, 2012 7:29 am Subject: Re: [phenixbb] Autosol: MIRAS

Hi AK, Did you do a wavelength scan when you collected the SE dataset you need to put the values of f' and f'' from your wavelength scan in order to locate the heavy atom sites, 6 methionine should be enough to phase your molecule. Shya

On Wed, Sep 5, 2012 at 9:25 PM, wrote:

...

Hi all,

I am trying to solve a structure through experimental phasing using AUTOSOL. I have a couple of heavy atom derivative datasets (Hg, La, Eu, Cd) and also a SeMet data. Unfortunately all the datasets are of low resolution (3.7-4.2A) and there are possibly 4-8 molecules in the asu. MIR, SAD and MAD alone did not give any convincing solution.

However, MIRAS, with a combination of few heavy atom datasets and the anomalous data from SeMet crystals, gave a somewhat satisfactory map. But the heavy atom site picked by AUTOSOL list only one of the heavy atoms i.e. Lanthanum. In another set of run, the solution of which was not convincing, the heavy atom substructure had only Hg. There are 6 Met out of 200 residues in one molecule and mass spec results show that Se incorporation is 100%.

Now, my doubt is that why does the heavy atom substructure contain only La and how can I get the substructure involving Se from this solution (or the datasets used)? Se location is going to help me a lot for finding a starting point to assign side chains.

Any suggestion would be greatly appreciated.

Thanks AK

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

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

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

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