User:Emw/CASP

CASP, which stands for Critical Assessment of Techniques for Protein Structure Prediction, is a community-wide experiment to compare and assess methods in protein structure prediction and related activities. Occurring biannually since 1994, the CASP experiment requires interested groups to predict the three-dimensional, tertiary structure of a set of proteins based on their amino acid sequence, or primary structure. The protein structures to be predicted are provided mostly by structural genomics centers (e.g. The Joint Center for Structural Genomics), who have determined the target's structure but agree to withhold publishing their data until the conclusion of the CASP experiment. Results and analysis for each CASP are published in special supplemental issues of the scientific journal Proteins. The best performing groups in CASP represent the state of the art in their particular areas of focus.

Prediction categories
Evaluation of results from both human and server groups in CASP is carried out in several prediction categories.

The foremost category of prediction is tertiary structure prediction, which is in turn divided into three subcategories based on how much information is available for the target protein at the time of prediction. Proteins with structural similar sequences (known as templates) available in the Protein Data Bank comprise the 'template-based modeling' (TBM) category, which comprises both the 'fold recognition' or 'new fold' categories from before CASP7. TBM predictions typically rely on homology modeling. The easiest, most accurate TBM targets form the 'high accuracy' (HA) category. The third category, 'free modeling' (FM), is composed of proteins for which no structural information is found and used.

Prediction of disordered regions in proteins has been done since CASP5.