Computational protocol: Molecular Evolution and Functional Divergence of Trace Amine–Associated Receptors

Similar protocols

Protocol publication

[…] HMMTOP (ver. 2.1) [] and Phobius (ver. 1.01) [] were used to predict the transmembrane protein topology, which includes N-terminal, transmembrane (TM), intercellular loop (IC), extracellular loop (EC), and C-terminal regions. [...] Selection patterns were tested using the maximum-likelihood framework developed by Goldman and Yang []. The site-, branch-, and branch-site models implemented in codeml of the PAML (Phylogenetic Analysis by Maximum Likelihood) package (version 4.5) were used []. We first used the site-model M0 (one-ratio, ω, for all sites) to estimate the dN/dS (ω) for each TAAR subfamily. Two sets of likelihood-ratio tests (LRTs; d.f. = 2) were performed for positive selection: M1a (two site-classes, nearly neutral model: 0 < ω0 < 1 and ω1 = 1) vs. M2a (three site-classes including positive selection: 0 < ω0 < 1, ω1 = 1, and ω2 > 1) and M7 (beta distribution and 0 < ω < 1) vs. M8 (beta distribution and ω > 1). Using the branch models, we performed LRTs with d.f. = 1 between a one-ratio model (R1; the same ω for all branches) and a two-ratio model (R2; two independent ω's) [, ]. As illustrated in , each test was set to compare primary amine-detecting TAAR lineages (TAAR1 and TAAR3) against tertiary amine-detecting receptor lineages (TAAR7 and TAAR8). We also used the branch-site models in order to detect positively selected sites along specific branches [, ]. In these models, positive selection was allowed on a specific, "foreground", branch, and the LRTs (d.f. = 1) were performed against null models that assume no positive selection. The branch-site test of positive selection ("Test 2" in []) has four site classes: 0, 1, 2a, and 2b. For the site classes 0 and 1, all codons are under purifying selection (0 < ω0 < 1) and under neutral evolution (ω1 = 1), respectively, on all branches. For the site classes 2a and 2b, positive selection is allowed on the foreground branches (ω2 ≥ 1) but the other, "background", branches are under purifying selection (0 < ω0 < 1) and under neutral evolution (ω1 = 1), respectively. For the null model, ω2 is fixed as 1. For our analysis, TAAR7 and TAAR8 subfamilies were tested. For each subfamily phylogeny, tests were done using each branch (from both internal and terminal branches) as the foreground. The numbers of tests performed were 61 and 26 for TAAR7 and TAAR8, respectively.All PAML analyses were carried out using the F3X4 model of codon frequency []. The level of significance (P) for the LRTs was estimated using a χ2 distribution with given degrees of freedom (d.f.) and the test statistic calculated as twice the difference of log-likelihood between the models (2∆lnL = 2[lnL1 –lnL0] where L1 and L0 are the likelihoods of the alternative and null models, respectively). Positively selected amino acid sites are identified based on Bayes Empirical Bayes posterior probabilities []. [...] Homology modeling of TAAR protein structures was performed using the SWISS-MODEL Web server (http://swissmodel.expasy.org) []. The same template, the B-chain of the turkey (Meleagris gallopavo) β1AR (4AMJ), was selected for the human TAAR1, elephant TAAR7a, and mouse TAAR8a proteins. See for the details on TAAR protein structural modeling. The graphical representation of TAAR structures was prepared with the PyMOL Molecular Graphics System (version 1.3, Schrödinger, LLC). […]

Pipeline specifications

Software tools PAML, SWISS-MODEL, PyMOL
Databases ExPASy
Applications Phylogenetics, Protein structure analysis
Chemicals Amines