PoDP RefSeq accession is used as antiSMASH accession

[CunliangGeng]

There are a few correlated issues:

1. PoDP schema shows that RefSeq_accession is will search on endpoint https://www.ncbi.nlm.nih.gov/nuccore/ (this is the endpoint of GenBank), but actually RefSeq has its own endpoint https://www.ncbi.nlm.nih.gov/refseq,

TODO:

• the RefSeq endpoint in the schema should be updated

---

2. PoDP users fill the RefSeq_accession section with Assembly accession, e.g. the project shown below, the GCF_000515175.1 is actually an Assembly accession but not a RefSeq accession
   

RefSeq and Assembly accessions have different prefixes:

• RefSeq accession prefix list
• Assembly Accession have two prefixes: GCF and GCA

  • The assembly accession starts with a three letter prefix, GCA for GenBank assemblies and GCF for RefSeq assemblies. This is followed by an underscore and 9 digits.

TODO:

• PoDP should validate the prefix to make sure the filled data is valid
• PoDP should add Assembly section for Assembly data
• PoDP should ask users to correct their mis-filling of RefSeq_accession section

---

3. (it looks) antiMASH database use Assembly accession as its own accession, e.g. antiSMASH GCF_000814765.1 and the corresponding Assembly data

When NPLinker tries to download antiSMASH data, it actually extracts the accession from PoDP RefSeq_accession section , and then use that accession to query antiSMASH database to download pre-run data. But just as discussed in question 2, RefSeq_accession is a wrong place to fill Assembly accession and PoDP schema should add another section for Assembly_accession.

Question

• does antiSMASH database always use Assembly accession as its own accession?

TODO:

• when PoDP is updated for Assembly accession, NPLinker should also be updated to extract accession from corrected PoDP section for downloading antiSMASH data

[justinjjvanderhooft]

Thanks for bringing this up @CunliangGeng! I have invited Marnix Medema to the NPLinker repo as well, as he is more versed in this area than I am, but I will do my best below.

[justinjjvanderhooft]

Re 3. Could you elaborate a little bit on what you mean here? I think in the future, it would be good to further improve the PoDP scheme and allow users to point to a (pre-run) antismashDB result - is that what you mean by "antiSMASH accession'? As I don't understand what that is....antiSMASH runs on sequenced genomes and results in predicted gene clusters for each genomic sample.

[justinjjvanderhooft]

Re 1. We are aware that the current scheme is not perfect and that sometimes genomes cannot be found. In a next version of the PoDP scheme, it could be worth checking if changing the endpoint results in a higher download % for most of the projects. Hard to say if this will the case based on our experience so far.

[justinjjvanderhooft]

Re 2. Let's discuss this in a meeting as things can get rather confusing rather quickly.... For example, for the project you are linking to, the genomic data is actually stored at JGI: i.e., https://img.jgi.doe.gov/cgi-bin/m/main.cgi?section=TaxonDetail&page=taxonDetail&taxon_oid=2515154093 - and when searching for the data at NCBI, you can find a suppressed record with the GCF code: https://www.ncbi.nlm.nih.gov/assembly/GCF_000515175.1/ - and there are others as well: https://www.ncbi.nlm.nih.gov/assembly/organism/168697/latest/

[CunliangGeng]

Hi Justin, I updated the issue description to try to make it more clear and answer your questions. Let's discuss this issue in next meeting.

[justinjjvanderhooft]

thanks for some clarifications - let's discuss further in our next meeting - also to what extent we can update the PoDP in the current project. Some lower-level adjustments are to improve/extent the user instructions :-)

[CunliangGeng]

Reported issues to PoDP iomega/paired-data-form#249.
Close this issue then.