SCit: Tools for protein side chain conformation analysis and editing

This server provides several tools for the analysis, quality assessment and editing of protein side chain conformations.

One feature of this series of tools is that is makes use of a Hidden Markov Model derived Structural Alphabet.
More information on HMM structural alphabets here.
More information on SA derived rotamers here.

SCit users guide:

• Load a PDB file for further processing.
  
• Automatically (re)build side chains for which some or all heavy atoms are missing
  Note that, at this stage, it simply consists in the building of side chains in a standard conformation. Further repositioning is possible using the next stage tools.
  
• Substitute side chains, if a sequence is specified
  
• (Re)position side chains corresponding to the uppercase letter of the sequence specified
  

The main page accepts as parameters:

1. PDB file: a PDB formatted protein structure. For RMN structures, only the first model is currently considered.
   The program checks the backbone geometry: incorrect geometry or backbone missing atoms may be not compatible with the HMM encoding. In such case, the program will stop. Another check is related to the presence of all side chain heavy atoms.
   
2. A sequence (Optional): The specified sequence length MUST exactly match that of the PDB file. It will be used to :
   Subsitute amino acids not matching the PDB file sequence.
   (re)Position the side chains. For such purpose, only parts of the sequence specified in upercase will be considered. Lowercase letters of the sequence will remain unaffected.
   

SCit second stage services:

• List Chi Values :

  This simply returns the values of the dihedral angles together with some information concerning their relative solvent accessible surface area (Residue exposed (>25% ASA) "E" or Residue buried (<25% ASA) "B") by residue.
  Example :
  
  
  

• Plot conformations :

  This tool will, for all or user selected amino acid types, produce plots of the conformations of the side chains of the protein (chi1 or chi1/chi2) superimposed onto density maps obtained from the analysis of a large collection of structures.
  
  For residue with only one chi (CYS, SER, THR, VAL), the plot consists in a histogram.
  Differents colors are used:
  
  
  Exposed residues (>25% ASA)
  Buried residues (<25% ASA)
  Residue for which the conformation is rather unusual with regards to those observed in a collection of structures
  (Regions corresponding to occurrence frequencies lower than a user specified threshold (0.005 at 0.040))
  
  
  Example of output:
  
  

• Edit Chis :

  The side chain conformation editing services allow to modify the values of the dihedral angles. Different methods are proposed:
  
  
  Automated side chain positioning:
  This tool positions the side chains in rotameric conformations. Different rotamer libraries are proposed.
  • The library denoted as "None" corresponds to a backbone independent Rotamer Library.
  • The library corresponding to "1 letter" is a backbone dependent rotamer library for which the dependence between backbone and side chain conformation is expressed as the dependence of the side chain conformation to the description of the local backbone conformation using one letter of the structural alphabet (i.e. a fragment of 4 residue length).
  • Lib2L is similar to lib1L but for words of two letters (i.e. fragments of 5 residue length).
  
  To position the side chains, we presently use a procedure similar to that described by Bower et al. (REF: Bower M.J., Cohen F.E. and Dunbrack R.L., Jr. (1997) J. Mol. Biol. 267:1268-1282), but using our backbone (in)dependent rotamer libraries. Compared to the original method, we use a simplified approach:
  1. First, we position each side chain in its most probable conformation having no clash with the backbone (i.e. we scroll the rotamers by decreasing occurrence frequency, and we stop as soon as the rotamer presents no clash with the protein backbone).
  2. Second, we detect the inter side chains clashes and, having sorted the side chains by decreasing clash order, we attempt to solve each problem in turn, continuing to scroll the list of conformations by decreasing occurrence frequencies.
  Such procedure is sub-optimal only, since we do not attempt to consider the combinations of side chain conformations.

  Example of output:
  
  
  Manual editing of side chain conformations:
  The modification of a chi value in the editor will result in the building of the required conformation.
  It is possible to :

  1. Change the diedral value in the editor for the selected residues.
     
     
     
     
  2. Change the conformation of all dihedral values at once by specifying one rotamer with r* (* is the number).
     
     
     
  
  Example of output:
  
  
  

• Substitute :

  This service allows to substitute side chains. This is achieved by editing the sequence displayed in the editor. You must use the 1 letter code for the 20 usual amino-acids. No insertion or deletion is authorised. The substitution of a residue by another type of residue will result in building a new structure. It is possible to attempt to somewhat optimize the substituted side chain conformations by selecting the "Automatically reposition substituted side chains ...".
  
  
  
  Example of output:
  
  
  
  

• Check Conformations :

  This tool will attempt to detect side chains in possibly incorrect conformations. This tool is based on the structural alphabet derived side chain conformation analysis. Besides the libraries, we have stored the densities of occurrence of the side chain conformations on grids of step 10 degrees. Analysing the variation of the densities observed for the side chain conformations in the different libraries, we have setup a procedure to assess if a given side chain conformation is likely to occur given one or two letters of the structural alphabet. For each type of side chain, given a set of predicted conformations, thresholds have been optimised on a collection of proteins to detect erroneous conformations.
  Note that (i) this does not corresponds to identify off-rotameric conformations, and (ii) that it is possible that, even for a Xray structure, some side chains are identified as having possibly incorrect conformations.
  More information on check side chain conformations: here.
  
  
  
  Risk level 1: side chains for which the conformation corresponds to a density more than a threshold in lib1L, and a density less than this threshold for lib2L (typical threshold values are on the order of 0.001)
  
  Risk level 2: conformations having densities less than the threshold for both lib1L and lib2L
  
  Example of output:
  
  
  

• List Rotamers :

  This tool will produce, for each residue, a listing of rotamers compatible with the structure.
  It is possible to produce a listing with each of the three rotamer libraries described above (one backbone independent Rotamer Library, and two backbone dependent rotamer libraries).
  
  
  Example of output:
  
  
  
  

• SC neighbours :

  This tool permits to identify the structural neighbours of the side chains. The search is based on the inter atomic distances between the atoms of the different side chains of the protein.
  Two side chains are identified in contact if they have at least one pair of atoms in contact (hydrogens considered).
  Two atoms are in contact if their inter-atomic distance minus the sum of their Van der Waals radii is less than the specified threshold.
  The identification of the structural neighbours can be performed:
  • on the basis of the current side chain conformations
  • on the basis of all their possible conformations. For this latter case, neighbours correspond to side chains for which for at least one pair of rotameric conformation, at least one pair of atoms is in contact.
  
  
  
  Example of output:
  
  

• Compare:

  This tool proposes to compare the current side chain conformations to those of a reference file. The amino acid sequence of the reference must be identical to that of the current file.
  
  
  
  Example of output: