Use case website web-enmr.cerm.unifi.it
In order to access the web site a personal certificate trusted by EGGEE installed in yours browser is needed.
If the certificate is not present in your browser the website will present only the link to The International Grid Trust Federation (https://www.eugridpma.org/members/worldmap/) where you can find information about how to obtain the certificate.
After this step, the site will check if you are registered to the Virtual Organization (VO) enmr.eu, if you are not registered the site presents you a link to the VOMS server (https://voms2.cnaf.infn.it:8443/voms/enmr.eu/Login.do) where you can fill the form for requesting the access to the enmr.eu VO .
After this procedure you can access your personal page on the web-site, your user will automatically created. The access to web site will be allowed from any Browser where your certificate are installed.
The site manages a Quota for disk space. To increase the amount of disk space send a request mail to admin enmr-grid@alpha.cerm.unifi.it.
To access any GRID resources, the first step is to create or check if exists a Proxy certificate. This can be done using the menu “Proxy-> Create Proxy” or “Proxy->Proxy Info”
After that you can start to run, check or retrieve your calculations.
Example of Paramagnetic Structure Calculation of a
Calbindin protein and refinement in Amber
The protein
calbindin D9k was extensively studied
from our group in order to test/apply the use of paramagnetism-based restraints
. The protein contains two diamagnetic calcium(II) ions, which can
alternatively be substituted with paramagnetic lanthanide(III) ions without
alteration of the protein structure.
To perform a Xplor-NIH structure calculation we have to start from the creation of a new project:
“Calculations -> Xplor Structure Calc -> new project “ create new-project and assign one name.
(ex. Calb_web) open the Calb_web project and press “new calculation”. A form appears and has to be filled with the input data and parameter of the calculation.
The input file can be downloaded from the address Calbindin_web.tgz Calbindin_web
Tar contain:
par_axis_3.pro parameter for the tensor axis
ca.par parameter for the Calcium Ion
protein.psf topology of the protein
CA2.psf topology of the Calcium Ion
axis_new_501.psf
axis_new_502.psf
axis_new_503.psf topology of the axis for the tensor
protein.pdb random initial structure of the Calbindin
ca.pdb Calcium Ion coordinates
axis_xyzo_3_501.pdb
axis_xyzo_3_502.pdb
axis_xyzo_3_503.pdb initial coordinates for the tensors
calbindin_NOE.tbl NOE constraints
metalcenter.tbl constraints that link Ca ion to the protein residues
tensors.tbl constraints that link tensors to Ca ion
calbindin_ACO.tbl Dihedral Constraints
calbindin_PCS_ce.tbl PCS calculated substitute Ce to Ca ion associated to axis_xyzo_3_501
calbindin_RDC_ce.tbl RDC calculated substitute Ce to Ca ion associated to axis_xyzo_3_501
calbindin_RDC_dy.tbl RDC calculated substitute Dy to Ca ion associated to axis_xyzo_3_503
The web form has to be filled with parameter, topology and coordinates files:
In the NOE form we have to insert the tensor.tbl alone and call this class “tensor”, normally we scale this by a value of 1000.
The NOE and metalcenter constraints can be put together in a single class, for example noe, and scale this by a value of 50.
Insert calbindin_ACO.tbl in Dihedral Constraint field.
PCS constraints:
Insert calbindin_PCS_ce.tbl file into the input field and add a name to the class, for example pcsce. In Tensor ax,Tensor rh insert the anisotropy value of the tensor, 521.4 and -175.75 respectively, fmed are not used for this calculation so put a value of 20, in the scale field put a value of 5.0.
RDC constraint:
For RDC constraints fill the field as described for PCS constraints following the next schema:
In metal parameter and metal topology you have to insert the Amber metal topology and parameter.
This field have to be filled following the same schema used in xlor parameter and topology file:
the residues name of the metal have to be called XM1 for the first metal and XM2 the seconds, XM3..etc.etc.
Following the next image schema we create an amber metal parameter file using a type atom name “M1”, the element Ca with Mass 40.080 and VdW parameter with r_min=2.1 and eps=0.05.
In the topology field we have to create the topology for the metal. The name of the residues has to be XM1 (as in xplor-nih topology) the atom name has to be CA2 (as in xplor-nih topology) and the type name (created in previous parameter fields) has to be M1. The total charge metal ion (2.0) and the field “num” has to be 1 (features implementation for generic residues)
In the last fields we can set the Simulated Annealing (SA) parameters:
Init T and High Step are the temperature and the number of step for the first step of SA, after that the system starts with 15000 Cooling Step.
Time step is the size of the integration timestep in this case 0.005 ps.
End Count is the number of structure calculated per job, Number of Threads is the number of job run on the selected Queue (Short) on the GRID.
In this example the total number of structures calculated will be 5x 10 -> 50.
After the submission the system start to send the jobs on the
GRID.
Click on the arrow to return to the project menu:
The Stop field open a windows where we can stop the jobs:
The Check Status open a windows where you can check the status of the calculation jobs in GRID:
When one or more jobs finish we can start the analysis of the results:
In the “Comp Sel” we can select the atom or residues (using
xplor syntax selection) to perform the RMSD, in the “Top Frac” we can select
the number of best structure ,ordered by Energies, used for RMSD (1 for all
structure and 0.2 for 20% of the best structures).
The structures will be ordered by Overall energy. Selecting “Download” you can download the single structure, with “Display” the structure will be displayed using Jmol plugin:
At this point we can select the structures to refine with Amber.
Depending by the system dimension select the appropriate queue (Short for calculation less then 2 hours , Medium less then 4 hours, Long less then 24 hours, Very long less then 72 hours), considering that short calculation have greater priority then Medium and so on.
The job that contain the calculation of all the structure selected will be created using the name of the father Xplor-NIH structure calculation with the amber suffix:
The job management will be the same as the previous structure calculation ones.
The “Analysis” link allow us to check and download all the resulting protein structures.
The link “Display” will shows the structure using Jmol applet, the “Download” link allows us to download the refined pdb structure, the “Violations” link shows the result of the “sviol” amber command that shows informations on the structure violations.