Ottoni 2021: 2 samples missing due to no exact sample ages

[ivelsko]

Bug description

The Ottoni 2019 dataset includes 2 samples that aren't in the host-associated metagenome table, although they are ancient dental calculus and I can't find any indication in the publication that they should be excluded from this table.

Samples:
ERS6256710 (3295-GR14)
ERS6256715 (3300-CT3)

Were these intentionally excluded, or should they be added?

Additional context

No response

[bbartholdy]

It looks like it was @jfy133 that added this study

[jfy133]

I don't remember excluding anything on purpose...

However that was a long time ago and the GitHub app won't let me see what the reviewer said for some reason 😅

If you don't see any reason to exclude them, I suggest making a PR to add the two missing samples and libraries :)

[ivelsko]

Oops, I had the wrong date. The correct publication is Ottoni2021, not Ottoni2019. I corrected the issue title

[alexhbnr]

I also did some digging. For these two samples we have sequencing data, however, James wasn't able to determine the approx. sample age from the information provided by the authors and therefore removed them (see 30ab5e0).

I double-checked and I cannot infer the sample ages either. For the Romanian sample GR14, the publication from 1976 doesn't seem to appear online and, for the Italian sample CT3, the publication that the authors cite only talks about everything to the Neolithic but not about the Bronze Age.

According to our sample column specification, we follow the guidelines that

⚠️ if a sample has no age, and cannot be inferred (sediment only) it cannot be included in AncientMetagenomeDir!

The question here is now whether it is more important to have a clean sample table or be inclusive regarding the sequencing data. @jfy133 @aidaanva

[jfy133]

Ah indeed, I think I forgot because most of that investigation was via private chat with the first author.

iMO I wanted a clean table at least with age, as that is a critical piece of info. Without which the 'ancient' definition is very questionable and not verifiable.

You could change this is you want but you may have to bqcl through all previous papers as there are a couple of others (only a handful mind) that have the same problem. I'm hindsight I should have probably recorded that, sorry about that

[alexhbnr]

OK, I will put it on the to do list to take a vote in the next AMDir core team meeting. If we agree to do it, it might be something for a next hackathon that we want to have at the end of the year.