Dynameomics Fragment Library

Our fragment libraries were created to identify, rank and retrieve protein structure fragments from both the Dynameomics data warehouse and Protein Data Bank (PDB) for protein loop structure prediction. Dynameomics fragment structures supplement crystal structure fragment conformations to improve loop predictions without being dependent on sequence homology. The fragment libraries are accessible either through our web interface or via web services. Each of these methods is described in detail below.

Web Interface

Our web interface was designed to generate backbone predictions for proteins from 3 to 22 residues in length. Queries can be performed on our fragment libraries generated from Dynameomics or PDB structures, containing approximately one billion fragment samples. Details of our library structure, search methods, and evaluations have been submitted for publication as referenced below.

The interface supports queries using PDB codes or custom PDB formatted text files. Queries can be performed on any one of the libraries individually or on a collection of libraries simultaneously. We also integrated our rotamer library into the search mechanism, so side-chain structures can be predicted along with the backbone structure by providing the appropriate sequence. Otherwise, backbone structures will be provided as all alanine residues. Select the link below to use the web interface.

Fragment Search Web Interface

Web Services

We also provide programmatic access to the fragment libraries using web services. A sample Microsoft .NET project is provided via the link below and includes a detailed example of using the fragment web services. The web service methods provide a minimum interface to our libraries by exposing the internal coordinates search method.

Fragment Web Service Example


Direct Web Service Access

If you have any questions regarding the DIVE software, email us at dynameomics_info@uw.edu.

Reference List

1. Rysavy SJ, Beck DAC, and Daggett V. Dynameomics: Data-driven Methods and Models for Utilizing Large-Scale Protein Structure Repositories for Improving Fragment-Based Loop Prediction. Submitted for publication, 2014.