In statistics cross-validation is the practice of partitioning a sample of data into subsamples such that analysis is initially performed on a single subsample, while further subsamples are retained "blind" in order for subsequent use in confirming and validating the initial analysis.
Cross-validation is important in guarding against testing hypotheses suggested by the data, especially where further samples are hazardous, costly or impossible (uncomfortable science) to collect. 
-- This quote is licensed under the GNU Free Documentation License. It uses material from the Wikipedia article "Cross_validation".
Cross validation is a model evaluation method. To evaluate a model, you need to predict on a part of your sample data that this model hasn't seen yet. Here are some methods:
it should be "ProteinS", dude Meng-Juei Hsieh and Ray Luo*, "A Physical Scoring Function Based on the AMBER Force Field and the Poisson-Boltzmann Implicit Solvent for Protein Structure Prediction", Protein, August 2004; 56(3): 475 - 486