[DataPLANT DNA extraction template] suggestions for entries in column "bio entity" [DBPO:0000012]

[UrsulaE]

As DNA extraction is such a central point in contemporary plant research, I would like to draw attention to an issue that has been bugging me for some time. Not sure here is the right place to mention it, though. Default DNA extraction methods extract both linear and circular DNA, yet the only suggested term including the term "DNA" is BAO:0000316 "genomic DNA" with a definition that excludes circular DNA. I suggest adding plain "DNA" (BAO:0000269) to the suggested terms. I also have a feeling that most suggested terms offered in the list below "protein" are not a suitable entries for "bio entity" (all these terms seem to come from the EFO and PR ontologies, if that was of help).

[Brilator]

The "bio entity" block was added to some templates – at a very early stage of swate template designs – to categorize, what is actually worked on / extracted (RNA, DNA, protein, metabolites). So "circular DNA" would also fit.

However, this is probably overall not quite "clean" as essentially it would be an attribute describing the output node

[StellaEggels]

Just to add to the issue what you mentioned yesterday: Unsuitable suggestions/child terms are also listed for the terms:

• organism (OBI): shows EFO; OBI and GO terms that should not be child terms
• growth facility
• plant structure development stage

[StellaEggels]

I now added DNA (BAO:0000269) to DPBO as a child term of bio entity.

[StellaEggels]

I think the issue is that not just child terms are shown when you double click in a cell, but also otherwise related terms / similar terms / xref terms. Maybe this can be changed in Swate @Freymaurer . Might have to be discussed, but I agree that showing only child terms would be better.

[Freymaurer]

Swate search was changed to provide maximum performance:

It now shows terms according to search and verifies terms if they are is-a related by the icon in front:

alternatively you can double click into the empty search field to trigger a "find all children (max 500)" search:

I hope this answers the question 🙂

[StellaEggels]

The problem is that for some terms (the ones mentioned above), the suggested terms should not be child terms:

[Freymaurer]

In some ontology they must be listed with is_a relationships towards that term?

[StellaEggels]

For growth facility it's true, they are listed in the MIAPPE ontology as child terms, maybe we can change that. But for the others I don't think that is the case. Bio entity is even a DPBO term, and shows a large number of EFO and PR terms as child terms.

[Freymaurer]

Some ontologies do cross reference foreign terms and simply add relationships to it. That might be an issue here.

[StellaEggels]

I think this is a Swate search issue, except for the MIAPPE ontology terms we were not able to find relationships that would explain why the terms are suggested. It seems like sometimes their parent terms are somewhat similar word wise (e.g. child terms of "collecting specimen from organism" are shown for "organism" or child terms of "plant structure" are shown for "plant structure development stage")

[Hannah-Doerpholz]

I had a deeper look into the issue for organism (OBI:0100026). Here, the listed child terms mainly include EFO, OBI and GO terms. The first thing I saw is that the exact same terms are suggested 4 times in the beginning. They are mainly from collecting specimen from organism (OBI:0600005, imported into EFO) and from OBI itself. Then, there is also MONDO:0009536 bacterial esophagitis and finally some NCBITaxon terms. While looking at the parent terms of some of the search results in both their respective ontology (EFO, OBI, GO, MONDO) I saw that one mostly common denominator is the term material entity BFO:0000040 . This appears in more than one of our ontologies (e.g. OBI, EFO, PATO). Depending on how the term is used in each ontology there might be more is_a properties that lead to the creation of this issue. Is it possible for someone with access to SwateDB to please look into BFO:0000040 in this regard? From what I can tell the problem might originate from somewhere in BFO, likely this specific term.

Additionally, for some reason NCIT:C14250 Organism, NFDI4PSO:0000030 Organism and MIAPPE:0041 Organism are all declared as parent terms of OBI:0100026 organism. In the DPBO file I see that these terms are all connected through xrefs. However, using the Swate API /api/IOntologyAPIv2/getAllTermsByChildTerm with this query

[
  {
    "Name": "organism",
    "TermAccession": "OBI:0100026"
  }
]

I still got those terms, even though they should be equivalent, not parents. This strange query might contribute to showing wrong suggestions in Swate.

[StellaEggels]

@Zerskk could you look into this?