AllosMod-FoXS Help Pages

The AllosMod-FoXS server is a combination of FoXS and AllosMod. AllosMod is a server to sample protein conformations, which was originally created to study allosteric proteins. FoXS is a server to do rigid body modeling of small angle X-ray scattering (SAXS) profiles. The combined server is meant to create a user friendly way to generate structures to compare to experimental SAXS profiles.

The structures that are generated are a result of an underlying energy landscape and sampling procedure. The energy landscapes are constructed with the assumption/hypothesis that the most stable structure for a given sequence is the native crystal structure. In some instances, there might be several stable structures, in which case the energy landscape should include multiple basins of attraction. AllosMod is a tool that allows the user to specify what structures should define the energy landscape. Crystal structures or models of protein, DNA, RNA, and sugars may be used.

There are different sampling options listed from least diverse sampling to most diverse sampling. As you go down the list, sampled structures will increasingly deviate from the inputted PDB files:

1) Generate static structures, using MODELLER, which are similar to the input structure

2) Generate models using fast AllosMod simulations at 300 K

3) Generate models using fast AllosMod simulations at temperatures between 300 and 500 K and with increased stability of secondary structure.

4) Generate models that strongly deviate from the crystal structure by sampling at high temperature and with chemical frustation (Weinkam et al. 2009 Biochemistry, 2394-2402). Chemical frustration allows for high conformational variability in regions of the structure with buried charges.

These default settings can be adjusted with more flexibility by using AllosMod, which allows the user to download the scripts and log files. These scripts can be ran on the user's computer using MODELLER. For help, click here. With AllosMod, you will also be able to generate models based on a thermodynamically well sampled energy landscape. If this energy landscape produces models that reproduce SAXS data, then perhaps the user can learn how the microscopic motions are relevant to the biological function of the protein.

Input your sequence as a string of capital letters, the same specifications used in MODELLER. Multiple chains can be specified by using a "/" as a separator. Support is also given for nucleic acids, cofactors, and small molecules (for more information, refer to the MODELLER library file or help pages). For help regarding specific input fields, refer to the AllosMod help page (for glycosylation, click here) and the FoXS help page.

HINT: Input your PDB structure into the first page and hit submit. Then copy the sequence for that file and hit the back button and paste the new sequence into the input field. Then you just need to make sure to add any HIS tags or other residues missing from the PDB structure. This procedure allows you to submit a job with minimal manual work.

An example input sequence is shown below. PDB code 3L2P contains one protein chain, three DNA chains, and one ligand (represented by a "."):



HMRHKDCLLREFRKLCAMVADNPSYNTKTQIIQDFLRKGGDVYLTVKLLLPGVIKTVYNLNDKQI
VKLFSRIFNCNPDDMARDLEQGDVSETIRVFFEQSKSFPPAAKSLLTIQEVDEFLLRLSKLTKED
EQQQALQDIASRCTANDLKCIIRLIKHDLKMNSGAKHVLDALDPNAYEAFKASRNLQDVVERVLH
NAQEVELSVQASLMTPVQPMLAEACKSVEYAMKKCPNGMFSEIKYDGERVQVHKNGDHFSYFSRS
LKPVLPHKVAHFKDYIPQAFPGGHSMILDSEVLLIDNKTGKPLPFGTLGVHKKAAFQDANVCLFV
FDCIYFNDVSLMDRPLCERRKFLHDNMVEIPNRIMFSEMKRVTKALDLADMITRVIQEGLEGLVL
KDVKGTYEPGKRHWLKVKKDYLNMADTADLVVLGAFYGQGSKGGMMSIFLMGCYDPGSQKWCTVT
KCAGGHDDATLARLQNELDMVKISKAAVWEITGAEFSKSEAHTADGISIRFPRCTRIRDDKDWKS
ATNLPQLKELYQLSK/jllletljltj/ltjllejtl/ljjeltjjlejlejljetjjjl/.