Figure 4

Checking the model fit: Histograms of the discrepancy statistics for the HNF4 α data. The red lines represent the value we observe from the data for the mean number of peaks per window, their standard deviation, and the maximum possible read score. The histograms are plots of the same variables estimated using the 3,000 simulated values of the parameters during the run of the algorithm. The closer the real value is to the simulated distribution, the better the model explains those aspects of the data. The first row shows the three histograms of the values generated using the Poisson model and the second row shows the corresponding values for the negative binomial model. The latter shows a notable improvement in explaining these features of the data.