Fig 1.
(A) Across-trait correlation and (B) partial correlation with a posterior median > 0.2 or < −0.2 (in color). HIV gag mutation names start with the wild type amino acid state, followed by the amino acid site number according to the HXB2 reference genome and end with the amino acid as a result of the mutation (‘X’ means a deletion). Country = sample region: 1 = South Africa, -1 = Botswana; RC = replicative capacity; VL = viral load; CD4 = CD4 cell count. (C) Conditional dependencies between HIV-1 immune escape mutations that affect RC or VL. Node and edge color indicates whether the dependence is positive (orange) or negative (blue).
Table 1.
Efficiency comparison among different sampling schemes (BPS, Zigzag-HMC, LG-HMC, LG-NUTS).
We calculate effective sample size (ESS) per hour run-time for the elements of C, R, X, log joint density log p(X, Ω), and likelihood l(X, Ω). For the three multivariate parameters (C, R, X) with dimensions 276, 276, and 11,235, respectively, we report the minimal ESS across all dimensions. We conduct three independent simulations for each method and report the ESS values in the first three rows. We include the mean and standard deviation in the last row for each method to provide a summary of its overall performance. The bold number indicates the highest value in each of the five columns. For BPS, given the larger number of iterations required to achieve convergence, we record one sample of X every 1,000 iterations to comply with storage limitations, and report upper bounds of the actual ESS by multiplying the ESS from thinned samples by 1,000.
Fig 2.
(A) Across-trait partial correlation among H1 glycosylation sites and host type with a posterior median > 0.2 or < −0.2 (in color and number). (B) HA structure of a 2009 H1N1 influenza virus (PDB entry 3LZG) with six glycosylation sites highlighted. Site 278 and 289 are in the stalk domain and all others are in the head domain. (C) The maximum clade credibility (MCC) tree with branches colored by the posterior median of the latent variable underlying H1 glycosylation site 289. The heatmap on the right indicates the host type of each taxon.
Fig 3.
(A) Across-trait partial correlation among N1 glycosylation sites and host type with a posterior median > 0.2 or < −0.2 (in color and number). (B)(C) The maximum clade credibility (MCC) tree with branches colored by the posterior median of the latent variable underlying N1 glycosylation site 44 and 68.
Fig 4.
Across-trait correlations (A) and partial correlations (B) with posterior medians > 0.2 or < −0.2 (in color).
BB = bumblebee.