Phylogenetics

A pathway that works in a pepper might work in a tomato. It might also fail in ways that are predictable from an afternoon of looking at sequence, which is considerably cheaper than finding out in a greenhouse eighteen months later. The usual way to start is to ask how similar the relevant genes already are across the organisms you are considering, and how they relate to each other.

That is what this engine is for. Alignment with MAFFT, tree inference on the result, and an interactive rendering through phylo.io. It will not tell you a cross-species design will succeed. It will tell you which candidates are close enough to be worth the attempt and which are far enough away that you should expect to do real work.

A gene, and a set of organisms. The catalog spans 65 of them, so for pathway genes the selection is usually a matter of ticking boxes rather than hunting for sequence — the capsaicinoid pathway enzymes in particular are already present across every host that carries them. Your own sequences can be added to the set, which is the common case when you are bringing in an organism the catalog does not cover.

Beyond that, the defaults are sensible and the knobs are there if you want them: nucleotide or protein alignment, the MAFFT strategy, and how the tree is built from the alignment. A first pass with everything left alone is usually enough to see the shape of the answer.

Alignment engine

MAFFT

Tree viewer

phylo.io, interactive

Input

Catalog genes across 65 organisms, or your own sequences

Modes

Nucleotide or protein

Returns

Alignment, tree, identity matrix

Engine class

CPU

Credit cost

Free

Typical runtime

Under a minute for a modest set; a few minutes for a large one

Free, because it is a CPU engine — and the fact that it is free changes how people use it. Nobody runs an exploratory alignment when it costs money. They run one when it costs nothing, and they run it five more times with different organism sets, which is where the useful observation usually turns up. Credits apply only to GPU engines, at $25 for 250 credits, one credit for roughly one Standard job.

Exports follow standard formats so the alignment and the tree can be taken into whatever tool the rest of your analysis lives in.

Early. This engine tends to sit at the front of a programme, before any constructs exist, when the decision on the table is which host to attempt. Researchers use it for exactly that. Ag-biotech teams use it to understand how far a design will travel across the cultivars they already grow. Educators use the tree as a teaching object, since a clickable phylogeny makes an argument about relatedness that a static figure cannot.

It also earns its place in reverse. When a construct behaves unexpectedly in a new host, going back to the alignment and finding that the catalytic region was never as conserved as everyone assumed is a common and unglamorous resolution. The 65 organisms in the catalog, spread across 70 series and six tiers from Pepper through Tier 5, exist because somebody asked this question first.