Figure 1From: On universal common ancestry, sequence similarity, and phylogenetic structure: the sins of P-values and the virtues of Bayesian evidenceExtreme value probability distribution for similarity scores. A standard extreme value distribution (EVD) is shown. In Karlin-Altschul statistics, similarity scores (S) for alignments of random sequences are assumed to follow an EVD. P-values are based on the tail probability, shown as the shaded blue area, corresponding to the probability of observing a similarity score greater than or equal to the observed similarity score (So) for a given alignment of interest.Back to article page