Probabilistic models of DNA copy number amplification. The models are component distributions of chromosome specific mixture models. Models are marked in the figure on separate lines. The probability of an amplification in each chromosome band is denoted using white to black scaling, where black indicates a chromosome band with a high probability of an amplification (p = 1) and white indicates a chromosome band with a low probability of an amplification (p = 0). Probabilities between 0 and 1 have been linearly scaled as shades of gray. Amplification patterns (based on the maximal frequent itemsets) are reported on the left side of the amplification models. Prevalence of the amplification patterns in terms of the number of cancer cases (Nc) is shown on the right side of the models.