Utilizes microbial genomic annotations - such as those output from prokka - to construct per-gene trees, which may then be used to perform Unweighted UniFrac calculations on a per-gene basis. We can perform either a single-gene analysis, yielding an effect size for each unique gene or a multi-gene analysis i.e. Meta UniFrac yielding an effect size for several groups of genes. Note that the latter does not scale well computationally and only small numbers of combinations are recommended at this point.
conda install -c bioconda bedtools hmmer fasttree unifrac -c anaconda scikit-bio
git clone https://github.com/ahdilmore/GenomicUF.git
cd GenomicUF
pip install -e .