WIP: Protect index

[percyfal]

Add protected keyword to index output.

[percyfal]

@LeandroRitter first and foremost check README. I suggest we wait with merge until we decide what to do with the reference sequence configuration.

[LeandroRitter]

Yes, agree, let me think and test it

[ZoePochon]

I think the documentation idea in this PR is still useful, but I would probably handle it as a small README update rather than merging the old code as-is. My understanding is that users do not strictly need to use the full NCBI nt or Microbial NCBI nt databases provided with aMeta, but because the workflow assumes NCBI-style taxonomic identifiers and compatible mappings between KrakenUniq, Bowtie2 and MALT, we should recommend that custom databases are built from a compatible subset of NCBI nt rather than from an unrelated taxonomy.

Suggested README text as a new FAQ point:

Can I use custom databases with aMeta?

Custom databases can in principle be used with aMeta, but KrakenUniq, Bowtie2 and MALT should use compatible reference sets and NCBI-style taxonomic identifiers. The databases do not strictly need to be the full NCBI nt or Microbial NCBI nt databases provided with aMeta, but the same taxa and sequence identifiers need to be traceable across the screening, mapping and authentication steps. We therefore recommend building custom databases from a consistent subset of NCBI nt. Otherwise, taxa detected during KrakenUniq screening may be absent from the Bowtie2 or MALT databases, which can prevent proper downstream validation or lead to misleading authentication results. Databases based on other taxonomic frameworks, such as GTDB, are not currently a documented or tested use case in aMeta. Users who wish to use such databases would need to provide compatible taxonomy, sequence-to-taxid and accession-to-taxid mappings across all relevant steps.