We will annotate KO's of the KEGG database using sourmash and Diamond, and then compare the results -- see how it goes.
Step 1: restrict the whole functional annotation pipeline within KEGG organisms. Step 2: run the existing pipeline at https://github.com/KoslickiLab/KEGG_sketching_annotation.
Our main focus, for now, is to implement Step 1.
conda create -n kegg_ko
conda install -y -n kegg_ko -c bioconda -c conda-forge -c conda --file requirements.txt
conda activate kegg_ko
For now, we have downloaded 114 genomes. Each of these have a fasta file, as well as a mapping file storing information of the genes present in that genome. Actual number of genes need to be determined and documented here.
Still not written, work in progress