Phylogenetics workflows

Builds a maximum-likelihood tree from homologous sequences with MAFFT, IQ-TREE 3 or FastTree, and sets out how far UFBoot, SH-aLRT and rooting can be read. Use when you need a reproducible gene tree with support values.

Databases & SQLPack: K-Dense scientific skillslinkphylogenetics
Author
Source
K-Dense-AI/scientific-agent-skills/skills/phylogenetics @ ec59a29
License
Unknown in the skill, MIT for the repository

About this skill

What it does

The skill takes homologous DNA or protein sequences to a tree with branch support. The agent aligns them with MAFFT (--auto or a named strategy such as L-INS-i or E-INS-i), trims with trimAl only when the alignment justifies it, then infers a maximum-likelihood tree with IQ-TREE 3 (ModelFinder -m MFP, 1000 UFBoot and SH-aLRT replicates, a fixed seed) or a quick approximate one with FastTree. Much of the text is about reading the result: UFBoot of 95 or more is a screening threshold, not proof; the 70% bootstrap rule does not apply to UFBoot; FastTree's SH-like support runs from 0 to 1; a reversible model does not find the root, so outgroup and midpoint rooting are display choices; a gene tree is not a species tree.

The work goes through scripts/phylogenetic_analysis.py. It validates the FASTA first — unique IDs without whitespace or Newick punctuation, the declared alphabet, no empty or ragged records — keeps IQ-TREE checkpoints unless you pass --redo, leaves earlier output intact when a tool fails, and summarises or draws the tree with ETE 3 while keeping IQ-TREE's slash-separated support labels verbatim. references/iqtree_inference.md covers partitioned multi-locus runs, concordance factors, ancestral states, LSD2 dating and codon models.

When to use it, and when not

Use it for a gene tree, a protein family or a microbial marker gene when you need a reproducible tree with model selection and support values.

The skill states plainly what the result does not settle: species trees, transmission direction, horizontal transfer, or dating from sampling dates alone. If the tree already exists and you only need to analyse it, see below.

How to run it

We do not copy this skill: it declares its licence as Unknown, which does not match the repository's MIT. Install it from K-Dense's repository with the button to the source: gh skill install K-Dense-AI/scientific-agent-skills phylogenetics for one skill, npx skills add K-Dense-AI/scientific-agent-skills for the set, or the whole repository as an Agent Plugins package for Cursor and Codex, where the skill may show under the scientific-agent-skills prefix. The tools are not installed for you: MAFFT and IQ-TREE 3 (or FastTree) must be on PATH, for example from bioconda, plus Python 3.12 with ete3==3.1.3 for summaries and PyQt5 for images; ETE 3 does not run on Python 3.13. Everything runs locally, with no API keys.

  • Claude Code: /phylogenetics, or ask for the skill by name.
  • Codex: $phylogenetics.
  • Cursor: type / in the Agent chat and pick phylogenetics.

Example: /phylogenetics Build an ML tree from rbcL.fasta (nucleotides), 1000 UFBoot, seed 42, and use Amborella as the outgroup in the figure.

How it differs from similar skills

ETE Toolkit for phylogenetic trees picks up where this one ends: a finished tree on ETE 4 — comparison, pruning, taxonomy, duplication and speciation events, figures. This skill uses ETE 3 only for a summary and a picture, and the two APIs are not interchangeable, so keep them in separate environments. Nothing else in our catalog builds phylogenies.

Install

This skill is not copied to the catalog. Install it from the author's repository.

Open the source

Connect through AnyModel API

This skill runs on whatever model your agent uses. Point Claude Code, Codex or Cursor at AnyModel to run it on Claude, GPT, Gemini and more with one key — pay per token, no subscription.

More from the pack K-Dense scientific skills