A complete molecular evolutionary analysis workbench covering the method space of PHYLIP, MEGA and PAUP*: pairwise and multiple alignment for nucleotides and amino acids, model-based distances, parsimony, maximum likelihood, ancestral reconstruction, codon models, motif discovery and epitope evolution, with publication-quality figures and tables.

At the centre is the Monte Carlo neutral-evolution null: observed distances and amino-acid substitution rates by physicochemical type are compared against what neutral nucleotide evolution would produce on the same tree, under a mutation process estimated from sites you nominate as neutral. No direction is assumed. A value below the envelope is evidence of purifying selection, above it of positive or diversifying selection, and both are reported on the same footing.

Each statistic carries its own measured properties: type I error under strict neutrality, and how far it drifts with the amount of divergence. Claims are promoted only where routes that do not share a tree agree on the direction.

Nine published datasets are bundled for learning, every one of them redistributable, alongside simulated sets whose true answer is known so you can check the methods against it.

Membership required. Every analysis runs in your own browser; no sequence you load is sent to the server.