Module FragmentMapper'

source code

Classify protein backbone structure according to Kolodny et al's fragment
libraries.

It can be regarded as a form of objective secondary structure classification.
Only fragments of length 5 or 7 are supported (ie. there is a 'central'
residue).

Full reference:

Kolodny R, Koehl P, Guibas L, Levitt M.
Small libraries of protein fragments model native protein structures accurately.
J Mol Biol. 2002 323(2):297-307.

The definition files of the fragments can be obtained from:

U{http://csb.stanford.edu/~rachel/fragments/}

You need these files to use this module.

The following example uses the library with 10 fragments of length 5.
The library files can be found in directory 'fragment_data'.

    >>> model = structure[0]
    >>> fm = FragmentMapper(model, lsize=10, flength=5, dir="fragment_data")
    >>> fragment = fm[residue]

Dice up a peptide in fragments of length "length".

@param pp: a list of residues (part of one peptide)
@type pp: [L{Residue}, L{Residue}, ...]

@param length: fragment length
@type length: int

Map all frgaments in flist to the closest
(in RMSD) fragment in reflist.

Returns a list of reflist indices.

@param flist: list of protein fragments
@type flist: [L{Fragment}, L{Fragment}, ...]

@param reflist: list of reference (ie. library) fragments
@type reflist: [L{Fragment}, L{Fragment}, ...]

Read a fragment spec file (available from
U{http://csb.stanford.edu/rachel/fragments/}
and return a list of Fragment objects.

@param size: number of fragments in the library
@type size: int

@param length: length of the fragments
@type length: int

@param dir: directory where the fragment spec files can be found
@type dir: string