Statistical models are relevant to many problems in computational biology, from gene and motif prediction to analysis of pathways and gene networks. Such models can describe observations associated with certain processes or phenomena. Learning these models from data can reveal interesting patterns and explain the observed properties of the system of interest. This is especially important for classification and prediction. Once a model architecture is chosen and a model is learned, it is typically used to assess whether a new instance is consistent with the source that generated the training data and is likely to be related to the same process.