README.md

Kraken on a BoaT

Retrieving busco data

We would like to fetch all the BTK buscos at some point, but for now, we will just extract a bunch from /lustre/scratch123/tol/teams/blaxter/users/sk13/minibtk_run/ directory, where are simply compressed.

echo -e /lustre/scratch123/tol/teams/blaxter/users/sk13/minibtk_run/*insecta* | tr ' ' '\n' | cut -f 10 -d '/' | cut -f 1,2 -d "_" > list_of_insects_with_busco

Which generated a list of 47 insect sample IDs.

sample IDs to BUSCO sequences

We want a .fa file with all the buscos for each species. Have made fetch_buscos.sh script for that.

for ID in $(cat data/busco_manual/list_of_insects_with_busco); do 
    bash ./scripts/fetch_busco.sh $ID;
done

Assembly ID to species

for ASMID in $(cat data/busco_manual/list_of_insects_with_busco); do 
    echo "$ASMID"
    echo -n "$ASMID, " >> data/busco_manual/list_of_species_with_taxonomy
    esearch -db assembly -query "$ASMID" | elink -target taxonomy | efetch -format native -mode xml | grep ScientificName | awk -F ">|<" 'BEGIN{ORS=", ";}{print $3;}END{print("\n")}' >> data/busco_manual/list_of_species_with_taxonomy;
done

This worked for most of them, but sometimes it just returns nothing (the API is not reliable). I rerun failed samples, for now, but in the future we should instead resolve taxonomy by querying GoaT

AMDID=GCA_927399515.1
curl "https://goat.genomehubs.org/api/v2/search?query=tax_lineage%28queryA.taxon_id%29&queryA=assembly--assembly_id%3D"$ASMID"&result=taxon&includeEstimates=true&summaryValues=count&ranks=&taxonomy=ncbi&offset=0&size=50&fields=none&names="

returns a JSON with the full lineage (needs a bit postprocessing, but works well).

exploring properties

TODO: load k-mers into python as ID -> set of k-mers TODO: find intersects of all the possible ID pairs

The point is to grasp some basic ideas about the dataset. How many species specific k-mers are there? How many universal?