Armadillin

This is an experimental tool under development. The recommended method for calling lineages remains normal Pangolin: https://github.com/cov-lineages/pangolin

A Re-engineered Method Allowing DetermInation of viraL LINeages

Armadillin is an experimental alternative approach to training models on lineages designated by the PANGO team.

Armadillin uses dense neural networks for assignment, which means it doesn't have to assume that positions with an N are the reference sequence. Armadillin is still very fast, in part because it sparsifies the feature input to this neural net during training.

Installation

With pipx (self-contained):

pip install --local pipx
pipx install  armadillin

or in your environment:

pip3 install armadillin

Usage

We'll use a NextStrain open sampled file for a demo:

wget https://data.nextstrain.org/files/ncov/open/global/sequences.fasta.xz

armadillin sequences.fasta.xz

or

armadillin sequences.fasta.xz > output.tsv

If you have sequences already aligned to the reference you can make inference much faster using the --seqs_are_aligned parameter:

wget https://data.nextstrain.org/files/ncov/open/global/aligned.fasta.xz
armadillin aligned.fasta.xz --seqs_are_aligned > output.tsv

Related tools

Pangolin is the OG for assigning lineages