search

petrelharp / ftprime_ms

4 stars 2 forks source link

Fig 5.A: meaning of horizontal lines #45

Closed jeromekelleher closed 6 years ago

jeromekelleher commented 6 years ago

I'm confused by the horizontal line in fig 5A between nodes B and C. Since E only inherits from B, should there not just be a direct line from B to E (like we have from B to A)?

I think we should have a quick explanation for the difference between horizontal and vertical lines. Or perhaps make one type dashed and the other full, to emphasise that they are qualitatively different?

petrelharp commented 6 years ago

I've made horizontal lines dotted and adjusted the caption. What do you think? I think that we should have the line in the pedigree between B and C, because those dotted lines mean "parents of", which is conceptually different than "contributed genome to".

jeromekelleher commented 6 years ago

This is good, thanks. I am still slightly confused by C being a parent of E, but not contributing any genome though... Am I just being too literal about the meaning of 'parent'?

petrelharp commented 6 years ago

Well, a frequent outcome of meiosis is transmission of one entire parental chromosome. Here, "parents" mean "the parental chromosomes that underwent meiosis to produce the gamete".

I should really remake this figure to show diploids, though: currently it is not a legal situation for diploids.

On Fri, Jan 5, 2018 at 2:49 AM, Jerome Kelleher notifications@github.com wrote:

This is good, thanks. I am still slightly confused by C being a parent of E, but not contributing any genome though... Am I just being too literal about the meaning of 'parent'?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/petrelharp/ftprime_ms/issues/45#issuecomment-355526457, or mute the thread https://github.com/notifications/unsubscribe-auth/AA_26Vlvlq2fOr7RT00ThyLVvKE3MBfnks5tHf5DgaJpZM4RTNmS .

petrelharp commented 6 years ago

I have re-made these figures: the resulting tree sequence has one more tree, but now this could be the result of diploids mating (and I have said so in the caption). I could cut out A if you think that would help?

The result also no longer has multiple chunks of ancestry coming together from different places into a single coalescent event at the top, which was a nice complicated bit about the previous example; but, simpler is better here.

jeromekelleher commented 6 years ago

Yes, I follow it now! Changing to explicitly talking about diploids makes this much easier to understand I think. This is a great figure; it really makes the forwards-time model totally clear.

Minor issues:

I haven't gone through the edges and trees in full detail yet, but I'll do it at some point.

petrelharp commented 6 years ago

whoops; yes. fixed.

ashander commented 6 years ago

This new fig looks great btw. Go diploids!

Also in #51 I took the time to copy/paste the tables into msprime and make it draw the trees. They agree =)