Responsive image

Getting Started


General description

LoopIng has been designed with the aim of providing candidate template structures for loop modelling. The method takes advantage of both sequence and geometry related features (e.g. loop sequence, sequence similarity, stem distance, stem secondary structure and stem geometry). These features are used in a Random Forest machine learning regression model that is trained to select the loop template with the lowest predicted 3D distance to the target loop among a list of putative ones. Given an input loop sequence, LoopIng find the set of candidate loops for the target loop, then uses the RF trained model to predict the RMSD distance between the target loop and each of the loops in a dataset of candidate loops. Candidate loops are ranked according to their predicted confidence level and the user can select the number of templates to be provided in the final output.

Guide to LoopIng home page

The web form for submitting a job is located in the body of the LoopIng homepage.

The Following fields are mandatory:
  • The PDB structure of the protein with the target loop to be predicted. The loop region can be either omitted from the PDB structure or included in it. In the last case the atomic coordinates of the target loop will be ignored. The input PDB structure needs to be uploaded using the “Load PDB File...” button. The allowed file extensions are .pdb, .res and .atm. In case a different file is provided, the file will not be uploaded.
    N.B.: The file must be in Protein Data Bank format (PDB) (see http://www.rcsb.org).
Responsive image


  • The amino acid sequence of the target loop to be modeled. Example: SDPVPEIEKPI
    N.B.: Loop length must be between 4 and 23 residues.
Responsive image


  • The position of the first loop residue (in terms of residue number). Example: 226
    The stems/anchors are defined as the main-chain atoms of the residues that precede and follow the loop. N.B.: Stems ("Loop start number-1 residue" and "Loop end number+1 residue") are preferred to have a regular secondary structure (alpha-helix or beta-sheet), otherwise a message will be provided in the results page.
Responsive image


  • The chain id of the loop (single letter). Example: A
    If no chain id is present the field has to be empty.
Responsive image


The following field is optional:
  • The number of desired output templates. Example: 5 (optional)
Responsive image

LoopIng result description

The output of the server is a table reporting information about the candidate template loops. The table has the following column headers:
  • Template PDB The PDB ID of the candidate template loop
  • Sequence The amino acid sequence of the candidate template loop
  • Chain The chain of the candidate template loop
  • Start The position of the first residue (in terms of residue number) of the candidate template loop
  • End The position of the end residue (in terms of residue number) of the candidate template loop
  • Resolution The resolution of the PDB structure of the candidate template loop
  • Stem Distance The stem distance of the candidate template loop (between the carbon alpha atoms of the residues preceding and following the loop region: "stems")
  • Confidence level The predicted confidence level of the candidate template loop
  • View By selecting the radio button displays the template loop structure in a Jmol window
  • Download Download the individual template loop structure (backbone only). The structure also includes the coordinates of two residues before and after the loop.

The following button allows to download a zip folder containing all the template loop structures:
Responsive image
Jmol window
The 3D structure of the selected template loop is displayed in a Jmol interactive panel. The stem distance is shown. The loop is shown in both lines and cartoon representation.
Please ensure that JAVA is installed and working on your computer and your browser configured accordingly. For any problem please refer to http://jmol.sourceforge.net