Single-chamber, air-cathode MFCs (130 mL) were constructed as previously described.
Phylogenetic trees based on maximum likelihood (ML) were constructed using RAxML (Stamatakis et al. 2005) with the PROTCATLG model and 100 bootstrap replicates.
While QSAR-ML is technology neutral, a reference implementation of tools to set up QSAR datasets complying with QSAR-ML was constructed as a set of plugins for Bioclipse [23].
The reference implementation of QSAR-ML is constructed as a set of plugins for Bioclipse and allows for graphical setup of datasets.
A ML phylogeny was constructed with RAXML v.7.0.3 using the rapid hill climbing bootstrap algorithm with 1000 replicates and following ML search.
The ML tree was constructed with PHYML v3.0 [33] under the LG substitution model [34] and the branch supports were assessed with 100 bootstrap replicates.
For the clade containing type II AFPs and fish skin mucus lectins, a separate ML tree was constructed using the WAG model, and a bootstrap test with 100 replications has been applied to test its robustness.
A maximum likelihood (ML) tree was constructed for each transmission pair, which included two unrelated subtype C reference sequences from the Los Alamos HIV Sequence Database as outgroup sequences to root the trees.
A new ML tree was constructed using the WAG model of amino acid substitution, followed by the parametric approximate likelihood ratio test (aLRT) to assess the significance of individual branches.
ML tree was constructed with RaxMl 8.0.17 [ 23].
