phylign

name	phylign
description	Phylign — BLAST-like alignment of genes, plasmids, or reads against phylogeny-compressed microbial genome collections (e.g. pre-2019 661k bacteria) on ordinary desktops. Uses phylogenetic compression (MiniPhy for building compressed assemblies/COBS indexes; Phylign for search via COBS + Minimap2). Use when searching huge genome warehouses without a cluster. Upstream: https://github.com/karel-brinda/phylign. Project: https://brinda.eu/mof. Route via homology-search; compare lexicmap.
license	MIT
category	analysis-tools
tags	["Phylign","MiniPhy","phylogenetic-compression","search","COBS","genomes"]
upstream	https://github.com/karel-brinda/phylign
stage	mapping

Phylign (phylogenetic compression search)

Upstream: karel-brinda/phylign · Compression builder: karel-brinda/miniphy · https://brinda.eu/mof

Citation

Břinda, K. et al. Efficient and robust search of microbial genomes via phylogenetic compression. Nat. Methods 22, 692–697 (2025). https://doi.org/10.1038/s41592-025-02625-2

See also docs/references.md.

Analytical thinking

Public microbial collections outgrow BLAST-style local search. Phylogenetic compression reorders genomes by estimated evolutionary history so XZ / COBS / Minimap2 work on highly compressible batches.

Component	Role
MiniPhy	Build phylogeny-compressed assemblies / indexes from genome batches
Phylign	Query compressed 661k-style DBs: COBS match → Minimap2 align

vs	Prefer Phylign when
lexicmap	Using phylogeny-compressed 661k / MOF collections on a laptop
mmseqs2	Nucleotide genome warehouse search (not protein catalogue)
minimap2 alone	Target set is millions of assemblies (needs compression/index stack)

Requirements (upstream): ~12 GB RAM, ~120 GB disk for DB + intermediates. Pin which compressed collection you downloaded.

How to run

# install deps (conda: python, snakemake, mamba) — see upstream README
git clone https://github.com/karel-brinda/phylign.git
cd phylign
make test
make download          # compressed assemblies + COBS indexes

# put/symlink queries, edit config, then:
make                   # or: make match && make map

Building new compressed libraries: use MiniPhy (https://github.com/karel-brinda/miniphy), then search with Phylign-style workflows.

Decision tree

Search vs huge bacterial genome collections?
├─ Phylogeny-compressed 661k / MOF on desktop → phylign
├─ Custom million-genome nucleotide index → lexicmap
├─ Protein homology → mmseqs2 / erast / foldseek
└─ Short reads → contigs for binning → strobealign / bowtie2

Related skills

homology-search · lexicmap · minimap2 · progenomes · gtdb · tool-selection