Between these two variables (Pearson correlation coefficient R = 0.64, R2 = 0.41, and p-value < 0.05, concluding that linear correlation different to 0 is statistically significant). (PDF) S6 Fig. Differential Distribution of Protein Phosphorylations per Clade. Boxplots showing the fractions of serine, threonine and tyrosine residues per protein per clade compared to the fractions of sites predicted to be phosphorylated by NetPhos [25] using a 0.75 cut-off. Significance analysis was carried out using non-parametric tests (Kruskal Wallis test for the comparison of 3 or more samples and Mann-Whitney U test with Bonferroni correction for the pairwise analysis). Differences in means are statistically significant (p-values << 0.05), except for the p53-p63 comparison of predicted phosphorylation factions (p-value = 1). (PDF) S7 Fig. Comparison of SEQ, SLT and PT with DOT rates. Combined profiles of normalized evolutionary rates per aligned site for family and clades (vertebrates set) comparing disorderorder transitions (DOT) with (A) amino acid substitutions (SEQ), (B) secondary structure elements-loop transitions (SLT), and (C) phosphorylation transitions (PT). Grey shaded areas delimitate Pfam domain regions. (PDF) S8 Fig. Comparison of SLT and PT with SEQ rates. Combined profiles of normalized evolutionary rates per aligned site for family and clades (vertebrates set) comparing amino acidPLOS ONE | DOI:10.1371/journal.pone.0151961 March 22,22 /Evolutionary Dynamics of Sequence, Structure, and Phosphorylation in the p53, p63, and p73 Paralogssubstitutions (SEQ) with (A) secondary structure elements-loop transitions (SLT) and (B) phosphorylation transitions (PT). Grey shaded 1471-2474-14-48 areas delimitate Pfam [66] domain regions. (PDF) S9 Fig. p53 clade in detail: graphical representation of sequence-based predictions. Heat map for structural traits plotted in the order of the p53 DNA-based phylogenetic tree get Elbasvir context, showing p53 protein names as boxes APTO-253 site colored according to the color guide in Fig 2. These heat maps are showing sequence-based predictions mapped to their corresponding residue sites in the multiple sequence alignment, after removing empty columns (i.e. columns fully gapped in the p53 clade) for this subset: (A) continuous structural disorder propensities by IUPred [15,62] with a color gradient from blue to white to red mirroring the disorder propensity gradient from low (blue) to high (red), with white being the boundary between order and disorder (remaining alignment gaps are colored fpsyg.2017.00209 in grey). (B) secondary structure predictions by PSIPRED [24,64] displaying 3 states loop (white), alpha helix (purple) and beta strand (yellow), and C) sites predicted to be phosphorylated by NetPhos [25] using a 0.75 cut-off (red). On top of these heat maps, normalized evolutionary rates per site are shown for amino acid sequence (SEQ) in green [26] vs. binary traits [27] of disorder-order transitions (DOT) in orange (upper left), secondary structure elements–loop transitions (SLT) in blue (upper center), and phosphorylation transitions (PT) in pink (upper right). All evolutionary rates were normalized with a mean of zero and standard deviation of 1: negative rates for slow evolving sites and positive rates for fast evolving sites. Grey shaded areas delimitate Pfam domain regions. (PDF) S1 Table. Accession numbers for the vertebrate datasets (i) and (ii). (PDF) S2 Table. PDB files and regions used for mapping DOT and disorder conservation into a structural c.Between these two variables (Pearson correlation coefficient R = 0.64, R2 = 0.41, and p-value < 0.05, concluding that linear correlation different to 0 is statistically significant). (PDF) S6 Fig. Differential Distribution of Protein Phosphorylations per Clade. Boxplots showing the fractions of serine, threonine and tyrosine residues per protein per clade compared to the fractions of sites predicted to be phosphorylated by NetPhos [25] using a 0.75 cut-off. Significance analysis was carried out using non-parametric tests (Kruskal Wallis test for the comparison of 3 or more samples and Mann-Whitney U test with Bonferroni correction for the pairwise analysis). Differences in means are statistically significant (p-values << 0.05), except for the p53-p63 comparison of predicted phosphorylation factions (p-value = 1). (PDF) S7 Fig. Comparison of SEQ, SLT and PT with DOT rates. Combined profiles of normalized evolutionary rates per aligned site for family and clades (vertebrates set) comparing disorderorder transitions (DOT) with (A) amino acid substitutions (SEQ), (B) secondary structure elements-loop transitions (SLT), and (C) phosphorylation transitions (PT). Grey shaded areas delimitate Pfam domain regions. (PDF) S8 Fig. Comparison of SLT and PT with SEQ rates. Combined profiles of normalized evolutionary rates per aligned site for family and clades (vertebrates set) comparing amino acidPLOS ONE | DOI:10.1371/journal.pone.0151961 March 22,22 /Evolutionary Dynamics of Sequence, Structure, and Phosphorylation in the p53, p63, and p73 Paralogssubstitutions (SEQ) with (A) secondary structure elements-loop transitions (SLT) and (B) phosphorylation transitions (PT). Grey shaded 1471-2474-14-48 areas delimitate Pfam [66] domain regions. (PDF) S9 Fig. p53 clade in detail: graphical representation of sequence-based predictions. Heat map for structural traits plotted in the order of the p53 DNA-based phylogenetic tree context, showing p53 protein names as boxes colored according to the color guide in Fig 2. These heat maps are showing sequence-based predictions mapped to their corresponding residue sites in the multiple sequence alignment, after removing empty columns (i.e. columns fully gapped in the p53 clade) for this subset: (A) continuous structural disorder propensities by IUPred [15,62] with a color gradient from blue to white to red mirroring the disorder propensity gradient from low (blue) to high (red), with white being the boundary between order and disorder (remaining alignment gaps are colored fpsyg.2017.00209 in grey). (B) secondary structure predictions by PSIPRED [24,64] displaying 3 states loop (white), alpha helix (purple) and beta strand (yellow), and C) sites predicted to be phosphorylated by NetPhos [25] using a 0.75 cut-off (red). On top of these heat maps, normalized evolutionary rates per site are shown for amino acid sequence (SEQ) in green [26] vs. binary traits [27] of disorder-order transitions (DOT) in orange (upper left), secondary structure elements–loop transitions (SLT) in blue (upper center), and phosphorylation transitions (PT) in pink (upper right). All evolutionary rates were normalized with a mean of zero and standard deviation of 1: negative rates for slow evolving sites and positive rates for fast evolving sites. Grey shaded areas delimitate Pfam domain regions. (PDF) S1 Table. Accession numbers for the vertebrate datasets (i) and (ii). (PDF) S2 Table. PDB files and regions used for mapping DOT and disorder conservation into a structural c.