Add more capture type options for janno file

[Tlkhi]

I would like to propose two additions to the janno file specification:

1. UDG_Note column

This column could be used to specify details about UDG treatment.

2. Archaic capture type

Currently, this capture type option is missing, so it should be added.

[Tlkhi]

Update:
ImmuneCapture,
WholeGenomeCapture &
ExomeCapture
Should be added as well

[stschiff]

Right, thanks for the input. Two quick replies:

On 1) We were thinking of a general system to encourage the use of *_Note fields, and would add a feature in Trident to sort columns in a way that *_Note fields would be listed right after the column they are referring to. Since we already allow arbitrary additional column, this could be a good way to add "soft" columns, so columns that aren't officially listed in the schema (which is getting bigger and bigger) but that are encouraged and treated with special treatments.

On 2) Right, we could add all those, but my worry is that they are very specific and not that often applied. More importantly, they actually change in nature, I think (other than the well-established Twist and 1240K captures), so it might then be unclear what exactly has been applied. The same is true for ImmuneCapture (I suppose multiple labs have different ones), WG capture (which I thought was only applied in one or two instances) or ExomeCapture (which is widely applied, but more in the medical community). But I'm certainly open for this.

[Tlkhi]

On 2) Right, we could add all those, but my worry is that they are very specific and not that often applied. More importantly, they actually change in nature, I think (other than the well-established Twist and 1240K captures), so it might then be unclear what exactly has been applied. The same is true for ImmuneCapture (I suppose multiple labs have different ones), WG capture (which I thought was only applied in one or two instances) or ExomeCapture (which is widely applied, but more in the medical community). But I'm certainly open for this.

There are samples with those Data Types already available tho, so I think it'd be good to have these data type options.

[stschiff]

Can you provide citations for these additional Capture Types, and note how many ancient DNA papers have used them? I still think that most of these would be custom capture reagents designed for one specific study (at least we have several of those ongoing in our institute), and so it wouldn't make huge sense to add these to the schema, in my view.

[Tlkhi]

Can you provide citations for these additional Capture Types, and note how many ancient DNA papers have used them? I still think that most of these would be custom capture reagents designed for one specific study (at least we have several of those ongoing in our institute), and so it wouldn't make huge sense to add these to the schema, in my view.

WholeGenomeCapture:
FregelPNAS2018
KrzewinskaSciAdv2018
McCollScience2018
Nieves-ColonMolBioEvo2020
SanchezPNAS2019
KilincCurrBio2016

ImmuneCapture:
BarqueraNature2024

ExomeCapture:
WangNature2021

[stschiff]

OK, so I'm open to adding these, in principle, but I just checked in Fregel 2018, Krzewsinska and Nieves-Colon that they are all referring to slightly different agents and procedures (I might be wrong), so I'm not sure whether this is really a consistent capture method.

I don't know, I guess there are two purposes of the column:

1. Just summarise the paper and report in the Janno what the authors write. In that case adding these additional types would make sense.
2. Using this for grouping and analysing data to avoid batch effects, in which case it would require high consistency in the lab procedures, I think. In this case I think it would not make too much sense to add categories which are themselves somewhat "buckets" for different protocols.

What do others think? Should we add ImmuneCapture. WholeGenomeCapture and ExomeCapture to the list of Capture_Types?

[Tlkhi]

OK, so I'm open to adding these, in principle, but I just checked in Fregel 2018, Krzewsinska and Nieves-Colon that they are all referring to slightly different agents and procedures (I might be wrong), so I'm not sure whether this is really a consistent capture method.

from KrzewinskaScienceAdvances2018
One individual, scy006, underwent whole-genome capture

from FregelPNAS2018 [supp notes]
Samples with low endogenous DNA content were enriched for human endogenous DNA using whole-genome in-solution capture (WISC)26. Briefly, human genomic DNA was used to prepare libraries with adapters containing T7 RNA polymerase promoters. These T7 libraries were submitted to in vitro transcription with biotinylated UTP to generate RNA baits covering the whole human genome

from Nieves-ColonMolecularBiologyandEvolution2020
35 samples were shotgun sequenced and 22 of these were subjected to whole genome enrichment (WGE)

[stschiff]

Krzewinska et al. 2018 quotes "commercially biotinylated probes from MYcroarray (www.mycroarray.com)"

Fregel et al. 2018 quotes "whole-genome in-solution capture (WISC)" and citing Carpenter et al. 2013 for the method.

Nieves-Colon et al. 2020 also cite Carpenter et al. 2013 in the Supplement (WISC), but additionally quote MYbaits Human Whole Genome Capture Kit (Arbor Biosciences).

Sanchez et al. 2019 also cites MYbait Human Whole Genome Capture, but there is no citation of Carpenter et al. 2018

Hmm... OK, so I think WISC after Carpenter et al. 2013 could indeed be easy to add, as that seems to be a reasonably well defined thing. We would then assume that all of these follow that. We could indeed just call it WholeGenomeCapture and make clear in the schema that it refers to Carpenter et al. 2013.

[nevrome]

Whole Genome Capture is a pretty unspecific term, unlike 1240K or Twist. I fear the exact meaning may very well change over time. We could call it WholeGenomeCaptureCarpenter2013 to preempt any confusion?

We have a definition of these types in the documentation: https://www.poseidon-adna.org/#/janno_details?id=library-properties. I think we should copy it to the schema.

[Tlkhi]

Any info on this one?
https://www.science.org/doi/10.1126/science.296.5566.261b

The authors from this paper have used it for some samples:
https://www.nature.com/articles/s41467-023-40198-w#Sec16

[stschiff]

This is about Cell Lines, not about a capture SNP panel. I agree with @nevrome's name suggestion and the suggestion to move definitions into the schema.