Back to "Biological Language Modeling Seminar Topics"

Back to "Protein structure prediction"

 

3D profiles

= method to encode primary amino acid sequence in a "structural" alphabet that is derived from pdb structure analysis:

Write a protein structure as a message in an alphabet of 18 letters, that are based on 18 classes, that are derived from analysis of protein structures with respect to the environment of each amino acid (J.U. Bowie, R. Luethy, D. Eisenberg):

1. main-chain hydrogen bonding interactions (i.e. secondary structure):

    helix

    sheet

    other

2. burial/surface exposure

    buried side-chain if accessible surface area is <40A^2

    partially buried if area is >40 and <114 A^2

    exposed if area is >114 A^2

3. polar/non-polar nature of its environment

==> 3 classes based on 1, and 6 classes based on 2+3

 

applies to family profiles only - why?

How well does this method work?

 

References:

1:  Rice DW, Fischer D, Weiss R, Eisenberg D. 
Fold assignments for amino acid sequences of the CASP2 experiment.
Proteins. 1997;Suppl 1:113-22.
PMID: 9485502 [PubMed - indexed for MEDLINE]

2:  Eisenberg D, Luthy R, Bowie JU. 
VERIFY3D: assessment of protein models with three-dimensional profiles.
Methods Enzymol. 1997;277:396-404. No abstract available.
PMID: 9379925 [PubMed - indexed for MEDLINE]

3:  Zhang KY, Eisenberg D. 
The three-dimensional profile method using residue preference as a continuous
function of residue environment.
Protein Sci. 1994 Apr;3(4):687-95.
PMID: 8003986 [PubMed - indexed for MEDLINE]

4:  Madej T, Mossing MC. 
Hamiltonians for protein tertiary structure prediction based on
three-dimensional environment principles.
J Mol Biol. 1993 Oct 5;233(3):480-7.
PMID: 7692069 [PubMed - indexed for MEDLINE]

5:  Luthy R, Bowie JU, Eisenberg D. 
Assessment of protein models with three-dimensional profiles.
Nature. 1992 Mar 5;356(6364):83-5.
PMID: 1538787 [PubMed - indexed for MEDLINE]

6:  Eisenberg D, Bowie JU, Luthy R, Choe S. 
Three-dimensional profiles for analysing protein sequence-structure
relationships.
Faraday Discuss. 1992;(93):25-34. Review.
PMID: 1290936 [PubMed - indexed for MEDLINE]

7:  Bowie JU, Luthy R, Eisenberg D. 
A method to identify protein sequences that fold into a known three-dimensional
structure.
Science. 1991 Jul 12;253(5016):164-70.
PMID: 1853201 [PubMed - indexed for MEDLINE]