Construct databases for all profilers from the same input refseq data - Githubissues

genomic-medicine-sweden / taxprofiler

Taxonomic profiling of shotgun metagenomic data

https://nf-co.re/taxprofiler

MIT License

0 stars 0 forks source link

Construct databases for all profilers from the same input refseq data #33

Open LilyAnderssonLee opened 10 months ago

LilyAnderssonLee commented 10 months ago

Construct databases using the same refseq data as the kraken2 database which includes

complete genome, chromosome, scaffold and contigs
Archaea, Bacteria, viral, protozoa, fungi, plasmid, human (GRCh38 &T2T), UniVec_Core

kraken2 database can be downloaded from https://benlangmead.github.io/aws-indexes/k2

The list of refseq sequences is stored in the tsv file: https://genome-idx.s3.amazonaws.com/kraken/pluspf_20231009/library_report.tsv

Update databases whenever the kraken2 database is updated.

This issue can be closed after construction of these databases:

[x] diamond
[x] kaiju
[ ] centrifuge
[ ] krakenuniq
[ ] kmcp
[ ] ganon
[ ] sourmash

LilyAnderssonLee commented 6 months ago

kaiju The db construction failed due the memory limit, 450GB on hasta. This database was built on UPPMAX Bianca with 512GB memory.