SenCS: generate Structure ENsembles of Sonformational States

General description and submission:

    Protein structures exist in an equilibrium of alternative conformations. SenCS utilzes the experimental data 
from NMR and X-ray crystallography and generates the structure ensemble that represent the dynamic properties of 
the protein under the experimental conditions. Briefly, it would process the NMR ensemble via distance thresholds
and perform time-averaged refinement for X-ray data.
    [the usage of the ensemble]
    To submit a job to SenCS, you need to input the PDB id of your interests. SenCS woud automatically determine the
type (NMR/X-ray) of the structure for sequential processing. For NMR structures, SenCS would finish the job in minutes.
As the time-averaged refinement for X-ray datasets generally takes hours, an email address is required to track the job. 
    [mention something about checking page?]

Funtionality of result interface

the NMR result interface

    The NMR module of the server retrieves the NMR ensemble from the PDB databse; it filters the residues and preserves the 
residues that fluctruate within a distance threshold (default 3.5A). 
    The result interface shows the filtered residues in meganta in the structure viewer. The structure viewer only loads the
first model of the ensemble for speed consideration. The animation that loads all models and visualizes the structure
dynamics is available via the button "Load animation". The plot 'Pairwise distance for residues' shows the maximal fluctruation 
for every residue with the range of pairwise distances. The filtered residues are shown in a separated data series.
    In addition to the default threshold, users can specify more distance threshold in the 'options' section ('generate new 
ensemble' option). All computed thresholds would be listed in the 'options' section, in which one can switch between different
thresholds and download the computed ensemble.

the X-ray result interface

    The X-ray module of the server retrieves X-ray dataset from the pdb_redo database and performed time-averaged refinement
for different pTLS values. 
    In the first section, it shows the name of the structure in PDB, as well as the Rfree and Rfactor values for the starting 
structure (usually the structure deposited in PDB) and the final ensemble. note that the R values for starting structure might 
be worse than the R values reported in the PDB database, probably because we have to filter out some atoms (usually from solvents)
that the program can not recognize and cause the program to terminate unexpectedly. 
    In the second section, the webpage shows the structure view of the protein colored by B-factors. For the speed consideration, 
the structure viewer only loads the first model of the ensemble, with the option to load all models in the ensembles and 
animate the fluctruation ('load animation' button). The plot 'Pairwise distance for residues' shows the maximal fluctruation
for every residue with the range of pairwise distances. 
    Depending on the options you input when submitting the job, the time-averaged refinement might be performed with a few pTLS 
values. All the available pTLS values, including the complete ones (black color), the computing ones (grey color with label 
'running'), and the aborted ones due to the insufficient number of included atoms (grey color with label 'aborted'), would be 
available in the 'option' section. Users can view the running log for all the available refinement runs, as well as switch the
interface and download the generated ensemble for the complete refinement runs. One can also perform additional refinement runs
with more pTLS values under the option 'Generate new ensemble'. An email address would be required to start the computation
of new ensemble because one refinement run would takes 2~7 hours.

Vocabulary

1. PDB id: the unique identifier of four letters in the PDB database

2. Time-averaged refinement: a protocal that performs MD simulations to refine a X-ray structure. specifically, we 
use the time-averaged refinement in phaser package (for more details, please read here)

3. Fast mode: run time-averaged refinement using all atoms for flexibility estimation (pTLS=1.0);

4. exhausive mode: run time-averaged refinement using an array of pTLS values (i.e. 1.0, 0.8, 0.6, 0.4, 0.2, 0.0) as the input.

5. email: a email address to track the job status.