Note: You may submit up to 5 jobs per hour.
Input consists of the amino acid sequence in FASTA or RAW format and the following two matrices:
- Dihedral angle matrix
- Contains the φ and ψ
dihedral angles of all protein
conformers. For N conformers of a protein with length
Np, this is a matrix of N rows and
2*(Np - 2) columns. Rows correspond to conformers;
columns to the angles φ2, ψ2,
φ3, ψ3, … ,
φ2 ψ2 φ3 ψ3 … φNp-1 ψNp-1 Conformer 1 Conformer 2 ⋮ Conformer N
- C-alpha coordinate matrix
- An N × 3Np matrix containing the
Cα atomic coordinates of all conformers.
Rows correspond to conformers (their order matching that in the dihedral
angle matrix). Columns correspond to the coordinates
x1, y1, z1,
x2, y2, z2, … ,
xNp, yNp, zNp,
where (xj, yj, zj)
is the position of the Cα atom of the jth
x1 y1 z1 x2 y2 z2 … xNp yNp zNp Conformer 1 Conformer 2 ⋮ Conformer N
These matrices must be submitted as whitespace-delimited, plain-text files. Fields should be floating-point numbers. To calculate the required dihedral angles from a set of PDB files, users may download the DSSP progam.
For convenience, users who have DSSP installed may download the shell script CollectProteinData.sh. This script assumes that PDB files for conformers reside in a common directory (for example, conformerDir) and have names of the form protein.n.pdb, where protein is a common prefix and n = 1, 2, … , N. To create the input files, run the following commands:
$ cd conformerDir
$ somePath/CollectProteinData.sh protein pathToDSSP
This will create the files protein.MainMatrix (the dihedral angle matrix), protein.Calpha (the Cα coordinate matrix), and protein.faa (the peptide sequence file) in conformerDir.
Output consists of atomic structures of the five top scoring conformers from each of the Cα-Cα and centroid-centroid distance dependent force fields. (Full structures are generated from Cα traces using the program REMO.) File names are of the form conf.field.rank.pdb, where conf is the row number of the conformer in the initial dihedral angle matrix, field is the force field employed (either calpha or centroid), and rank is a number between 1 and 5—with 1 corresponding to the lowest energy structure as determined by the given force field. Results are sent to the email address provided above.
Reports or publications of results obtained via the ICON server should acknowledge its use by citing the following:
Subramani, A.; DiMaggio, P. A.; Floudas, C. A.
Selecting High Quality Protein Structures from Diverse Conformational Ensembles.
Biophys. J. 2009, 97, 728-1736.