Figure 1
From: Intraspecies characterization of bacteria via evolutionary modeling of protein domains
Fit of protein domains RSA. (a) Example of protein domains Preston plot fitted with three different distributions: the Poisson Log-Normal, the Negative Binomial and the Log-Series. Results refer to the bacterial genome \(\text {GCA}\_000717515\). The Negative Binomial and the Log-Series fit overlap. This implies that the dispersion parameter \(\alpha\) of the Negative Binomial distribution (see Eq. (6)) is close to zero. The mean and the median of the dispersion parameter obtained for the 3368 bacterial genomes are \({2.67\times 10^{-4}}\) and \({2.62\times 10^{-7}}\), in agreement with the observed overlap. (b) Distribution of the difference between the AIC obtained with the Poisson Log-Normal model (PL) and the Log-Series (LS) or the Negative Binomial (NB) model, considering all the 3368 bacterial genomes.
