Abstract
Chemera is a molecular modelling software package that includes the algorithms BiGGER (Bimolecular complex Generation with Global Evaluation and Ranking), for modelling protein interactions and protein complex structures [1,2,3], and PSICO (Processing Structural Information with Constraint programming and Optimisation), to integrate experimental and theoretical data to solve protein structures [4, 5]. This paper focuses on the constraint programming aspects of Chemera, namely constrained docking, which allows the user to restrict the search for protein-protein complex models in a manner consistent with the ambiguity of some experimental data, and the processing of structural constraints to generate approximate models of protein structures from heterogeneous data. This allows the user to take advantage of experimental data obtained by a wide range of techniques, from spectroscopy to site-directed mutagenesis, and to integrate these data with theoretical considerations such as homology models, secondary structure prediction, or reaction mechanisms. For modelling protein complexes, this is done by specifying sets of potential contacts between the two proteins and how many of those contacts must be enforced, without having to specify exactly which contacts to enforce, which models the possibility of some of experimental results being due to effects other than proximity to the docking partner. For modelling the structure of a single protein, the information can be encoded as distances between atoms, as fixed relative positions for groups of atoms, or of constraints on angles of rotation around atomic bonds connecting rigid groups.

URL:
http://rewerse.net/publications/rewerse-publications.html#REWERSE-RP-2006-164

BibTeX:

@inproceedings{REWERSE-RP-2006-164,
	author = {Ludwig Krippahl and Pedro Barahona},
	title = {Chemera: Constraints in Protein Structural Problems},
	booktitle = {Proceedings of Workshop on Constraint Based Methods for Bioinformatics, Nantes, France (25th September 2006)},
	year = {2006},
	pages = {30--45},
	url = {http://rewerse.net/publications/rewerse-publications.html#REWERSE-RP-2006-164}
}