Related Experiment Videos
EVA: Evaluation of protein structure prediction servers
Ingrid Y Y Koh1, Volker A Eyrich, Marc A Marti-Renom
1Columbia University Center for Computational Biology and Bioinformatics (C2B2), Russ Berrie Pavilion, 1150 St Nicholas Avenue, New York, NY 10032, USA. koh@cubic.bioc.columbia.edu
Nucleic Acids Research
|June 26, 2003
Summary
EVA, a weekly updated web server, evaluates automated protein structure prediction methods. It provides reliable comparisons of secondary structure, contact, and modeling predictions using extensive data.
Area of Science:
- Computational Biology
- Structural Bioinformatics
- Bioinformatics
Background:
- Automated protein structure prediction is crucial for understanding protein function.
- Numerous prediction servers exist, with methods constantly evolving.
- Accurate evaluation of these methods is essential for scientific progress.
Purpose of the Study:
- To introduce and describe EVA (Evaluation of WWW Protein Model Servers), a web server for assessing automated protein structure prediction accuracy.
- To provide a reliable and continuously updated platform for comparing various prediction methodologies.
Main Methods:
- EVA automatically collects daily predictions from various servers using new Protein Data Bank (PDB) sequences.
- Predictions are compared weekly against experimentally determined protein structures.
- Results are published on the EVA web pages, accumulating a large dataset over time.
Main Results:
- EVA assesses secondary structure prediction, contact prediction, comparative modeling, and threading/fold recognition.
- The server handles a large volume of predictions, ensuring reliable method comparisons.
- Accumulated data spans hundreds to thousands of proteins, depending on the prediction method.
Conclusions:
- EVA offers a valuable resource for developers and users of protein structure prediction tools.
- The automated, weekly updates ensure EVA remains current with evolving prediction methods.
- Reliable and extensive evaluation by EVA aids in advancing the field of protein structure prediction.