Closed kitchensjn closed 1 year ago
Good question; under the hood this is basically handled via different flags to simplify( )
; SLiM's job is to record everything-ever and delegates "what's left in the tree sequence" to calls to simplify.
So, I think what you're asking for is a flag to simplify that says "please keep also recombination nodes"? And, I think you're right, it's more of a tskit discussion.
Yes, this is a tskit issue I think? @petrelharp how do we move the issue over there? Or perhaps @kitchensjn should just file a new issue over there, and then I'll close this one?
I'll repost on tskit. Thank you!
What would be the best way to retrieve the full ARG from a SLiM simulation tree sequence output? With
initializeTreeSeq(retainCoalescentOnly=F)
, we get a tree sequence with lots of unary nodes, which is necessary to include the recombination nodes. Is there a method for specifically marking these recombination nodes within SLiM versus searching for these afterwards usingtskit
? A full ARG tree sequence could be a helpful optional output, where recombination nodes are the only unary nodes retained. Additionally from what I can tell, nodes in the tree sequences from a SLiM simulation can have multiple types. One example being a sample node or coalescent node can also be a recombination node which would require two separate node flags in the nodes table; I don’t think this occurs withmsprime
. Do you have guidance for how to handle this (may be better suited intskit
Discussion)?Copying @pderaje as he is working on this with me. Below is our SLiM simulation: