MOOSE

Modified Organism Screening by Enrichment

Prototyping plan

Loosely based on the Debode paper:

1. Quality trimming (fastp)
2. Align reads to reference genomes of big 6 (bwa)
3. Extract RPKM values for each host
4. reference alignement to elements (SAUTE)
5. Find elements in contigs (BLAST)
6. Count reads in contigs ang get coverage
7. Try to detect event specific sequences (JRC-EURL Methods) in contigs

TODO

• Create reports
• Check if kraken instead of bwa doable
• chop contigs and reblast part with no matches? set a min length to chop (100bp?) use full BLAST db (local) hit selection? (best hit?)
• BLASTGraph improvement, check out ggcontigs (R)

Ideas for future dev

• Compare bwa to simple kraken?
• If bwa upgrade to bwa2
• Automated event detection (API to BCH/Euginius)?

Notes

BWA reference indexing:

1-Get assemblies from e.g. Genbank (wget or use ncbi browser)

2- concatenate asemblies:

cd path/to/assemblies
cat *.fna > merged_refs.fa

3- Index genomes

bwa index merged_refs.fa

Then use merged_refs.fa as a reference

Panel sequences

Multifasta format, unwrapped sequences!