(e.g. alignment site 471) are, in general, expected to conserve kinase

(e.g. alignment site 471) are, in general, expected to conserve kinase partner and thus, conserve the regulatory mechanism, while transitions from Ser/Thr phosphorylations to Tyr phosphorylations suggest divergent mechanisms of regulation via different kinases. Alignment site 164 in p63 clade switches from Ser in ray-finned fish to Tyr in the rest of species (with shark as an exception), suggesting divergent regulation in ray-finned fish p63 proteins. Thus, differential regulation within orthologs is implied. Also, alignment site 165 is known to be LY2510924MedChemExpress LY2510924 phosphorylated in human p63 (Tyr36) in PhosphoSite, but this phosphorylation site is missing in all fish, where shark has Cys and the others Phe or Leu.DiscussionUsing linear sequence predictors, properties of structural disorder (IUPred), secondary structure (PSIPRED), and phosphorylation sites (NetPhos) have been inferred. It is important to remember that these are predictions and cannot be perfect given that they are (i) independently aiming to predict traits that may depend on each other, (ii) using only the linear sequence context without considering long-range sequence contacts, and (iii) based on experimental data that may not reflect the dynamic nature of a protein sequence, e.g. one PDB structure is merely a snapshot of a conformational ensemble [36]. The accuracy for PSIPRED is >80 compared to actual experimentally determined protein structures [37]. For disordered proteins, fewer proteins are experimentally determined to be disordered. For IUPred, comparing to IDEAL (a small database of disordered proteins [number of proteins = 207]) [38] the accuracy is approximately 85 , but comparing to DisProt (a slightly larger database of disordered proteins [number of proteins = 794]) [39] the accuracy is approximately 62 SART.S23506 [40]. However, it has been found that IUPred is more accurate in predicting order vs. disorder for DisProt proteins if thePLOS ONE | DOI:10.1371/BX795 site journal.pone.0151961 March 22,14 /Evolutionary Dynamics of Sequence, Structure, and Phosphorylation in the p53, p63, and p73 ParalogsFig 7. Shared and clade-specific predicted phosphorylation patterns. (A) WebLogos [35] per clade showing 66 alignment positions following a 50 majority rule of phosphorylation predictions based on a phosphorylation score cut-off = 0.75 (NetPhos), gaps included. (B) Phosphorylation predictions mapped onto their alignment sites (numeration based on the full alignment), with scores ranging from 0 j.jebo.2013.04.005 (blue) to 1 (red) with 0.5 as the midpoint (white). Gaps are shown in grey. The colored boxes on the left show the distribution of species sorted by the phylogenetic tree following the color scheme as in Fig 2. Shared and clade-specific phosphorylation sites are distributed along domains (yellow shaded areas) and linkers. Sites marked with a circle means p53 clade-specific (black, the phosphorylation site is experimentally validated in PhosphoSite; grey, an adjacent site is experimentally validated to be phosphorylated in PhosphoSite). Sites marked with a star are predicted to be phosphorylated in a p63 or p73 clade-specific manner while p53 has a different experimentally verified posttranslational modification [34]. doi:10.1371/journal.pone.0151961.gcut-off is set to 0.4 instead of the intended 0.5 [39,41]. In a different study, IUPred predictions of 0.4 were frequently found for disordered residues in partially disordered proteins [42]). Thus, we used the 0.4 cut-off to infer order vs. disorder. The sensiti.(e.g. alignment site 471) are, in general, expected to conserve kinase partner and thus, conserve the regulatory mechanism, while transitions from Ser/Thr phosphorylations to Tyr phosphorylations suggest divergent mechanisms of regulation via different kinases. Alignment site 164 in p63 clade switches from Ser in ray-finned fish to Tyr in the rest of species (with shark as an exception), suggesting divergent regulation in ray-finned fish p63 proteins. Thus, differential regulation within orthologs is implied. Also, alignment site 165 is known to be phosphorylated in human p63 (Tyr36) in PhosphoSite, but this phosphorylation site is missing in all fish, where shark has Cys and the others Phe or Leu.DiscussionUsing linear sequence predictors, properties of structural disorder (IUPred), secondary structure (PSIPRED), and phosphorylation sites (NetPhos) have been inferred. It is important to remember that these are predictions and cannot be perfect given that they are (i) independently aiming to predict traits that may depend on each other, (ii) using only the linear sequence context without considering long-range sequence contacts, and (iii) based on experimental data that may not reflect the dynamic nature of a protein sequence, e.g. one PDB structure is merely a snapshot of a conformational ensemble [36]. The accuracy for PSIPRED is >80 compared to actual experimentally determined protein structures [37]. For disordered proteins, fewer proteins are experimentally determined to be disordered. For IUPred, comparing to IDEAL (a small database of disordered proteins [number of proteins = 207]) [38] the accuracy is approximately 85 , but comparing to DisProt (a slightly larger database of disordered proteins [number of proteins = 794]) [39] the accuracy is approximately 62 SART.S23506 [40]. However, it has been found that IUPred is more accurate in predicting order vs. disorder for DisProt proteins if thePLOS ONE | DOI:10.1371/journal.pone.0151961 March 22,14 /Evolutionary Dynamics of Sequence, Structure, and Phosphorylation in the p53, p63, and p73 ParalogsFig 7. Shared and clade-specific predicted phosphorylation patterns. (A) WebLogos [35] per clade showing 66 alignment positions following a 50 majority rule of phosphorylation predictions based on a phosphorylation score cut-off = 0.75 (NetPhos), gaps included. (B) Phosphorylation predictions mapped onto their alignment sites (numeration based on the full alignment), with scores ranging from 0 j.jebo.2013.04.005 (blue) to 1 (red) with 0.5 as the midpoint (white). Gaps are shown in grey. The colored boxes on the left show the distribution of species sorted by the phylogenetic tree following the color scheme as in Fig 2. Shared and clade-specific phosphorylation sites are distributed along domains (yellow shaded areas) and linkers. Sites marked with a circle means p53 clade-specific (black, the phosphorylation site is experimentally validated in PhosphoSite; grey, an adjacent site is experimentally validated to be phosphorylated in PhosphoSite). Sites marked with a star are predicted to be phosphorylated in a p63 or p73 clade-specific manner while p53 has a different experimentally verified posttranslational modification [34]. doi:10.1371/journal.pone.0151961.gcut-off is set to 0.4 instead of the intended 0.5 [39,41]. In a different study, IUPred predictions of 0.4 were frequently found for disordered residues in partially disordered proteins [42]). Thus, we used the 0.4 cut-off to infer order vs. disorder. The sensiti.