SAXS

Rigid Body Modelling :: identical subunits in P1?

ewa — Tue, 07 Sep 2010 09:05:42 GMT

Author: ewa
Subject: identical subunits in P1?
Posted: 2010.09.07 10:05 (GMT 1)

Have I understood it correctly that two identical subunits cannot be modeled in P1 using BUNCH? I have a dimer of two identical monomers, but not symmetry related, according to relevant crystal structures.

Mixtures and Flexible Systems :: GAJOE terminates prematurely

Alexandra — Tue, 07 Sep 2010 02:15:09 GMT

Author: Alexandra
Subject: GAJOE terminates prematurely
Posted: 2010.09.07 03:15 (GMT 1)

Hi:

I have run Ranch and generated pdbs (all in the same directory as the *dat file), now I am trying to run GAJOE on Linux using the default setting, but here is the error I get:

CYCLE: 49
Initial genes created
Starting Genetic Algorithm
CYCLE: 50
Initial genes created
Starting Genetic Algorithm
list in: end of file
apparent state: internal I/O
lately reading direct formatted internal IO
No such file or directory
list in: end of file
apparent state: internal I/O
lately reading direct formatted internal IO
No such file or directory
list in: end of file
apparent state: internal I/O
lately reading direct formatted internal IO
No such file or directory
list in: end of file
apparent state: internal I/O
lately reading direct formatted internal IO
No such file or directory
list in: end of file
apparent state: internal I/O
lately reading direct formatted internal IO
No such file or directory
mv: cannot stat `04557.pdb': No such file or directory
No such file or directory
list in: end of file
apparent state: internal I/O
lately reading direct formatted internal IO
No such file or directory
list in: end of file
apparent state: internal I/O
lately reading direct formatted internal IO
No such file or directory
list in: end of file
apparent state: internal I/O
lately reading direct formatted internal IO
No such file or directory
mv: cannot stat `05688.pdb': No such file or directory
No such file or directory
list in: end of file
apparent state: internal I/O
lately reading direct formatted internal IO
No such file or directory
list in: end of file
apparent state: internal I/O
lately reading direct formatted internal IO
No such file or directory
list in: end of file
apparent state: internal I/O
lately reading direct formatted internal IO
No such file or directory
mv: cannot stat `06780.pdb': No such file or directory
No such file or directory
list in: end of file
apparent state: internal I/O
lately reading direct formatted internal IO
No such file or directory
mv: cannot stat `07831.pdb': No such file or directory
No such file or directory
list in: end of file
apparent state: internal I/O
lately reading direct formatted internal IO
No such file or directory
list in: end of file
apparent state: internal I/O
lately reading direct formatted internal IO
No such file or directory
list in: end of file
apparent state: internal I/O
lately reading direct formatted internal IO
No such file or directory
Experimental file Zero_alexpool_B2and3and5

It looks like to program cannot files the files? Is there a specific naming/directory structure one needs?

Thanks,
Alex

Ab Initio Methods :: DAMCLUST

Alexandra — Sun, 05 Sep 2010 13:11:41 GMT

Author: Alexandra
Subject: DAMCLUST
Posted: 2010.09.05 14:11 (GMT 1)

IS there a reference for DAMCLUST? I would like to know more about how the clustering is done and the use of such clustering algorithms.

For example, let's say one does multiple simulations for a multidomain protein. For all the reconstructions, the core seems to align well, but one extension seems to adopt multiple oritentiations in the resulting ab initio models. Would you conclude this may correspond to an multiple conformations in solution?

Thanks!

Mixtures and Flexible Systems :: Refining dummy atom positions

sdw — Wed, 01 Sep 2010 12:28:14 GMT

Author: sdw
Subject: Refining dummy atom positions
Posted: 2010.09.01 13:28 (GMT 1)

Hi, I have recently collected scattering curves of a protein that I believe is forming homo-tetramers. I believe this because we have collected SAXS data for a truncated version of the protein, for which have tetrameric crystal structure, which gives a excellent chi*2 value in crysol. This same model fits the SAXS data for the full-length protein very poorly in part, I assume, because the individual monomers are missing approximately 50 residues off their N-termini and 10 residues of the C-termini. The caveat here is that we expect these missing sequences to be flexible and unstructured.
Obviously I would like to use EOM to build an ensemble of chains that would best fit the data, but if I understand correctly EOM needs (assumes) that you have a monomer of linked structured subunits, rather than an oligomer. As an alternative I have crudely used Ranch to generate a pdb file with random coil extensions for each chain. (Taking the eomtmp.pdb file generated), and then stitched these models together to generate a very crude model. This dummy atom extended model gives a better chi*2 fit to my SAXS data (much better than the truncated crystallographic model as well as a CHADD build on this), but could clearly be improved. Therefore I'm wondering if this model can be fed into any ATSAS programme to refine the dummy atom positions against the scattering curve (or more likely the P(r) plot) whilst maintaining correct connectivity of said dummy atoms and avoiding obvious steric clashes?

(Apologies for the long winded question)

SAS Software and Instrumentation :: RE: Problem with Primus

konarev — Wed, 01 Sep 2010 07:39:41 GMT

Author: konarev
Subject: Problem with Primus
Posted: 2010.09.01 08:39 (GMT 1)

Hi Arnab,

the current version of PRIMUS is compiled for Windows,
if you are using windows emulators in Linux/Mac, it can have some
problems.

In some cases it is possible to continue to work with PRIMUS
even if it gives an error message after starting, you need to press
'Ok' button and this message will disappear.
You can also try to run PRIMUS from earlier ATSAS2.2 version,
may be on your machine it will work better.

We have also plans to release a cross-platform version of PRIMUS in
next ATSAS release (in October-November 2010).

Peter

SAS Software and Instrumentation :: RE: Problem with Primus

arn007c — Tue, 31 Aug 2010 18:09:13 GMT

Author: arn007c
Posted: 2010.08.31 19:09 (GMT 1)

Hi Peter,

Thank you for replying. The problem is it is giving the error before it asks for any input. I was able to work with gnom from the same atsas package which is accepting the input data file in the recommended format. But the primus is not working. It might be a broken file or something I guess. Is there any way to check? Also is there atsas or similar program for mac?

Thanks,

Arnab

Data Processing :: RE: troubles with primus crashing in windows 7- 64 bit

konarev — Tue, 31 Aug 2010 16:27:48 GMT

Author: konarev
Subject: primus
Posted: 2010.08.31 17:27 (GMT 1)

Hi

we made some tests for PRIMUS on Windows 7 64-bit machines,
and it worked stable, so may be you have some processor-specific
problem or problem connected with your data formats.

The current version of PRIMUS (in ATSAS 2.3) is compiled for Windows
platforms and therefore can have some conflicts when running using
windows emulators from Mac/Linux.

For next ATSAS releases (2.4 etc) we are planning to make a cross-platform
version of PRIMUS and SASPLOT.

Cheers,
Peter

Mixtures and Flexible Systems :: RE: svdplot

konarev — Tue, 31 Aug 2010 16:17:16 GMT

Author: konarev
Subject: svdplot
Posted: 2010.08.31 17:17 (GMT 1)

Hi Terry,

if you run SVDPLOT from PRIMUS then you can only see the results
and the plots of singular vectors in the window.
To get the files of singular values and singular vectors, one needs
to run SVDPLOT (or SVDPLOTQW) separetely in the command line:
"svdplot datafile.dat /K",
where datafile.dat should contain all experimental curves
There will be three output files:
_U_.dat - with singular values
_S_.dat - with singular vectors.
_Ncomp.log - with the information for the number of components
from the statistical test.

Peter

Mixtures and Flexible Systems :: RE: Adjustment in Oligomer?

konarev — Tue, 31 Aug 2010 16:01:58 GMT

Author: konarev
Subject: Adjustment in Oligomer?
Posted: 2010.08.31 17:01 (GMT 1)

Hi Annette,

we have already discussed this problem in e-mails, but I want to summarize the important points on the forum as well.

If you are using OLIGOMER in a standard mode, then the fitted curve should be equal to a "pure" linear combination
from the form-factor components with the corresponding coefficinets (volume fractions) multipled by the scaling coeffifient.

It is also possible to use a key "/cst" (when you run OLIGOMER in batch mode), then additional (constant) component will be added (it can be used when you have poor fit of the experimental curve at higher angles)
but this constant component is choosen in such a way that its volume fraction is equal or very close to zero, so you should get the meaningful volume fractions for "real" components and improved fit at higher angles.

"Adjust" in PRIMUS finds the scaling and constant coefficients between two curves using least square method (the constant can be positive or negative), whereas in OLIGOMER the system of linear equations is being solved using non-negativity condition for the volume fractions.

Peter

Mixtures and Flexible Systems :: RE: Mixture run

konarev — Tue, 31 Aug 2010 15:54:26 GMT

Author: konarev
Posted: 2010.08.31 16:54 (GMT 1)

Hi,

please try to run the graphic version of mixture (mixtureqw.exe) from PRIMUS or use mixture.exe version from ATSAS2.2.
It seems the command-line version of MIXTURE in ATSAS2.3
is not running properly on some processors, it will be fixed
in next ATSAS release.

Rigid Body Modelling :: RE: Merge and Maths

konarev — Tue, 31 Aug 2010 15:48:51 GMT

Author: konarev
Posted: 2010.08.31 16:48 (GMT 1)

Hi

the data files are merged using a least-squares fitting in the overlapping range. For this, the experimental data are rebinned onto a new angular master grid and, if requested, the merged curve is recomputed on this grid. This option is useful e.g. for making composite curves from the data
recorded in different angular ranges. For all operations, the error
propagation is performed using standard equations.
Another approach is to join the points in the overlapped region, it works
well if you have data on the same grid, this approach was used in the
older PRIMUS versions (see ATSAS 2.1), in ATSAS 2.2 and 2.3
the rebinning method is used.

SAS Software and Instrumentation :: RE: Problem with Primus

konarev — Tue, 31 Aug 2010 15:41:32 GMT

Author: konarev
Subject: Re: Problem with Primus
Posted: 2010.08.31 16:41 (GMT 1)

Hi Arnab,

do you mean, that you could not load your data file
in PRIMUS ? Please, check the correct input format for
data files: it should be in ASCII format and contain
two or three columns: the first column - s-vector, the second column -
intensity, the third column (if exists) - errors of intensity.

Peter

Announcements :: RE: ATSAS version 2.3.1 released

rdobson — Tue, 31 Aug 2010 02:30:06 GMT

Author: rdobson
Subject: autoporod
Posted: 2010.08.31 03:30 (GMT 1)

Hi All,

I wondered how autoporod decided where to cut the data. Does it do it based on the error of the profile?? Actually, is there a rule of thumb regarding how to truncate noisy data??

Cheers
Ren

Rigid Body Modelling :: RE: Slit smearing in SASREF and BUNCH

david.a.jacques — Mon, 30 Aug 2010 22:03:45 GMT

Author: david.a.jacques
Posted: 2010.08.30 23:03 (GMT 1)

How does this work without asking for a beam profile, or some description of beam geometry?

PDB Oriented Tools :: RE: Prediction of scattering from dimers in Crysol

SaxsMax — Mon, 30 Aug 2010 15:45:14 GMT

Author: SaxsMax
Posted: 2010.08.30 16:45 (GMT 1)

In Crysol, the scattering is not normalized by the concentration
and I(0)~(delta_rho*Volume)^2~MW^2,
I.e. for the dimer, the forward scattering would be 4 times larger compared to monomer, if the hydration shell would not be taken into account (you could easily check this by running Crysol with zero contrast of the shell).

As you mention yourself, the shell of the two cubes side by side is different
from the two shells of the two individual cubes and therefore the ratio of four
is hold only approximately for the forward scattering intensities of a dimer and a monomer computed by Crysol

Ab Initio Methods :: RE: choosing the correct model's symmetry

SaxsMax — Mon, 30 Aug 2010 15:13:15 GMT

Author: SaxsMax
Posted: 2010.08.30 16:13 (GMT 1)

Judging from NSD values those mostly around 1.0 (except those with P222
symmetry), the models should be similar.

Mixtures and Flexible Systems :: RE: unit of q values in oligomer

SaxsMax — Mon, 30 Aug 2010 15:09:08 GMT

Author: SaxsMax
Posted: 2010.08.30 16:09 (GMT 1)

If you have the smax value of 5 in your data, then it is most probably in nm^-1 and you need to use /un 2
But the formfactors have to be computed up to 0.5 (A^-1)

Mixtures and Flexible Systems :: RE: limits of bunch and eom

SaxsMax — Mon, 30 Aug 2010 15:04:20 GMT

Author: SaxsMax
Posted: 2010.08.30 16:04 (GMT 1)

Why not start it once interactively and note all the answers in an answer file?

PDB Oriented Tools :: RE: Prediction of scattering from dimers in Crysol

tankozavr — Mon, 30 Aug 2010 14:59:36 GMT

Author: tankozavr
Posted: 2010.08.30 15:59 (GMT 1)

Hi! Thanks for your reply. I would like to clarify this further if possible.
Mw of the dimer is obviously twice of that of the monomer, concentration of the dimerized protein should be 2 times lower. The surface of the dimer cannot be that easily determined, I believe it depends on the dimer nature. (just by stucking 2 spheres or 2 cubes we are getting smaller surface and smaller surface/mass ratio, but we do get larger suface itself for a single particle) I would initially think that hydration shell of a dimer will depend on the protein nature and surface area, is this correct?

My sincere apologies if my questions and reasoning are quite simple or silly but I am wondering where am I wrong in it.

Thanks !

Mixtures and Flexible Systems :: RE: Residue limit in RanCh

SaxsMax — Mon, 30 Aug 2010 14:51:01 GMT

Author: SaxsMax
Posted: 2010.08.30 15:51 (GMT 1)

Please try to split the lines in the sequence file

Ab Initio Methods :: RE: dammif with very large complex

SaxsMax — Mon, 30 Aug 2010 14:49:03 GMT

Author: SaxsMax
Posted: 2010.08.30 15:49 (GMT 1)

Normally, the data must be cut up to the first good data point as defined
by AutoRg. If this results in the warning in Gnom (Dmax x Smin > Pi),
than the Dmax and the models should be treated with caution.
In general, if you have a protein-DNA complex it is better to use Monsa (given that the scattering curve from DNA alone or protein alone is available),
as Dammin is applicable to uniform, density particles only.
But for 5% of DNA it is probably still OK.

Ab Initio Methods :: RE: chi square, Rvalue, fit

SaxsMax — Mon, 30 Aug 2010 14:42:28 GMT

Author: SaxsMax
Posted: 2010.08.30 15:42 (GMT 1)

R-value concerns the regularized smooth data
whereas chi operates with the raw experimental data.
Therefore Chi is the decisive parameter.

Literature :: RE: Reference for ATSAS online

SaxsMax — Mon, 30 Aug 2010 14:40:06 GMT

Author: SaxsMax
Posted: 2010.08.30 15:40 (GMT 1)

This one may be relevant

Petoukhov, M.V., Konarev, P.V., Kikhney A.G. & Svergun, D.I. (2007) ATSAS 2.1 - towards automated and web-supported small-angle scattering data analysis. J. Appl. Cryst. 40, s223-s228

Ab Initio Methods :: RE: MONSA - average jobs

SaxsMax — Mon, 30 Aug 2010 14:33:59 GMT

Author: SaxsMax
Posted: 2010.08.30 15:33 (GMT 1)

You could try the modified version of damaver at
http://www.embl-hamburg.de/~maxim/addon/monsaver.zip
which is however not an "official" part of Atsas package

Rigid Body Modelling :: RE: asymmetric homodimer

SaxsMax — Mon, 30 Aug 2010 14:15:55 GMT

Author: SaxsMax
Posted: 2010.08.30 15:15 (GMT 1)

It is not very clear if you finally used the symmetry constraint or not.
By design, Bunch is only able to handle the cases with one chain per
asymmetric part, i.e. the dimer must be run with P2 symmetry which imposes that the two-fold axes of N- and C- termini coincide.
We plan to distribute a new program in the next release of Atsas
which would be able to handle complicated cases like yours.
One more thing about Bunch: the length of the linkers must be
at least four DRs otherwise some domains may stuck though they
were not fixed.