Phylogenetic tests of positive selection for target proteases.

−2ΔL, likelihood ratio test to detect positive selection; p: proportion of sites under positive selection for M8 model; ωa: dN/dS estimate for M0; ωb: dN/dS estimate for M8 model; cPositively selected sites identified by M8 model: amino acid sites found to be under positive selection with posterior probabilities greater than 90% (blank), 95% (underlined) or 99% (bold) in the BEB analysis. a The reference sequence is human SERPINB3. NA, not applicable because the neutural model fits better than positive selection.
** Significance with p<0.001.
* Significance with p<0.05.