search

solgenomics / sgn

The code behind the Sol Genomics Network, Cassavabase and other Breedbase websites
https://solgenomics.net
MIT License
66 stars 35 forks source link

Define breeding stage for accessions #3144

Open lukasmueller opened 3 years ago

lukasmueller commented 3 years ago

Expected Behavior

Options should include "hybrid", "inbred", "clone", "backcross", "mixed", "f2", "f3", "f4"...

Add these types as "accession_type" stockprop and display statistics on accession_type use on trial detail page. Would be listed under attributes in the stock detail page and could be queried for stats on trial overview pages.

For Bugs:

Environment

Steps to Reproduce

bauchetg commented 3 years ago

If I may chime in :-) Such option/notation of "generation stage" is to reflect a certain degree of heterozygosity and/or the association selection stage.

For example, a user self annotate a line as an f2 wheat line harvested year 2020 under a XYZ name. Could the same stock still be named XYZ in the next cycle and now being an f3?

A solution may be to append the "generation stage" to stock XYZ associated seedlots rather than to the actual XYZ accession stockbe updated for example?

For example, corn hybrid breeding runs side by side an inbred and hybrid programs. Keeping in mind the XYZ accession can be used in two different programs, are we on the same stock ID or shall we think of seedlots again?

I guess what I am trying to say :-) is the generation stage annotation, requires a good compromise between user self annotation and some automatic quality control/incrementation procedure within breedbase. A further implementation could be automatic incrementation #2866

bauchetg commented 3 years ago

An alternate path may be to assign the "generation stage" prop to the corresponding trial plot and keep the stock away (this is less constraining on seedlot use for example).

In that case, it could be interesting to set a relationship between a plot testing f4 in Trial A and a plot testing the following f5 in Trial B, so that stages can be queried in a logical manner without interfering with genealogy/seedlots? Would that keep generation stage info apart between genealogical and experimental data?

Sorry if I am thinking out loud :-D

lukasmueller commented 3 years ago

The generation stage is definitely a property of the seedlot and not the accession... Maybe we need easier ways to generate this information associated with seedlots?

lukasmueller commented 3 years ago

Bryan: generation of the clone is a property of the plot, it should be stored in the plot stockprop (Titima), if no seedlots are available

bellerbrock commented 3 years ago

@bauchetg assigning a "generation stage" prop to the corresponding trial plot sounds like a great solution to me. Would also be applicable for tracking clonal "generation" in sweetpotato (i.e. how many years the clonal material used in a given plot has been in the field building up virus pressure)

chris263 commented 1 year ago

How is it going? I think we should reopen this discussion.