is it an ARG?

[petrelharp]

In the latest commit, 764c0e2cf904d50407f449d8bbf609c7e8a48a60, I changed the section talking about the ARG. Better?

I still kinda want to call the first diagram in Fig 3 an ARG, since that's pretty much exactly what it is, I think? But since we're moving away from saying "ARG" elsewhere as shorthand for pedigree plus recombination, there's no need to say it there.

@jeromekelleher @molpopgen @ashander

[jeromekelleher]

I really like it anyway, thanks @petrelharp!

Figure 3 isn't quite an ARG I think under the strict definitions (we'd have to have events as the nodes for that), but I think it's fine to call it an ARG here for the sake of clarity. It's a good diagram showing how we disentangle the history of the sample and simplify it down to the interesting bits.

Looking at it again though, I get the familiar ARG-diagram headache trying to figure out what happened... What do the horizontal lines between (e.g.) B and C mean?

[petrelharp]

That's standard pedigree notation for mating. Horizontal line connecting mates, with vertical lines going down to offspring. Then I think of the ARG as labeling those lines according to which part of the genome goes left and which goes right.

[jeromekelleher]

Ahhhhh! Well, forgive me for saying this, but wouldn't it be clearer to call this an "annotated pedigree" then? By calling it an ARG, I was looking for RE and CA event nodes---I honestly don't see how the term is helpful here.

Also, the result page returned by the link above made me laugh out loud. I can't help feeling there was a coded message in there somewhere!

[petrelharp]

Heh. =)

On Tue, Oct 24, 2017 at 1:14 AM, Jerome Kelleher notifications@github.com wrote:

Ahhhhh! Well, forgive me for saying this, but wouldn't it be clearer to call this an "annotated pedigree" then? By calling it an ARG, I was looking for RE and CA event nodes---I honestly don't see how the term is helpful here.

Also, the result page returned by the link above made me laugh out loud. I can't help feeling there was a coded message in there somewhere!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/petrelharp/ftprime_ms/issues/8#issuecomment-338910750, or mute the thread https://github.com/notifications/unsubscribe-auth/AA_26cfD1LLXgLTTDq6XBzt0YrfpZis1ks5svZxegaJpZM4QDXe3 .

[molpopgen]

I like the term "annotated pedigree".

On Tue, Oct 24, 2017 at 1:14 AM Jerome Kelleher notifications@github.com wrote:

Ahhhhh! Well, forgive me for saying this, but wouldn't it be clearer to call this an "annotated pedigree" then? By calling it an ARG, I was looking for RE and CA event nodes---I honestly don't see how the term is helpful here.

Also, the result page returned by the link above made me laugh out loud. I can't help feeling there was a coded message in there somewhere!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/petrelharp/ftprime_ms/issues/8#issuecomment-338910750, or mute the thread https://github.com/notifications/unsubscribe-auth/AGHnH71yibwUEe5SFAN_CrAqgjXef1Vbks5svZxegaJpZM4QDXe3 .

--

Kevin Thornton

Associate Professor

Ecology and Evolutionary Biology

UC Irvine

http://www.molpopgen.org

http://github.com/ThorntonLab

http://github.com/molpopgen

[petrelharp]

I like the term "annotated pedigree".

I think to most people it would mean a pedigree annotated with disease status. How about "haplotype pedigree"? Or, "haplogree"? "recombigree"? (j/k)

[jeromekelleher]

I think to most people it would mean a pedigree annotated with diseasestatus. How about "haplotype pedigree"? Or, "haplogree"? "recombigree"?

Haplotype pedigree seems fine to me, but nowhere near as great as recombigree!

[molpopgen]

"The recombigree records the process of recombigration."

This will work out well...

On Tue, Oct 24, 2017 at 11:50 AM Jerome Kelleher notifications@github.com wrote:

I think to most people it would mean a pedigree annotated with diseasestatus. How about "haplotype pedigree"? Or, "haplogree"? "recombigree"?

Haplotype pedigree seems fine to me, but nowhere near as great as recombigree!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/petrelharp/ftprime_ms/issues/8#issuecomment-339091903, or mute the thread https://github.com/notifications/unsubscribe-auth/AGHnH90fUSqk02J7mdc9WtaQyYoXKA3Tks5svjBYgaJpZM4QDXe3 .

--

Kevin Thornton

Associate Professor

Ecology and Evolutionary Biology

UC Irvine

http://www.molpopgen.org

http://github.com/ThorntonLab

http://github.com/molpopgen

[petrelharp]

I think to most people it would mean a pedigree annotated with disease status. How about "haplotype pedigree"? Or, "haplogree"? "recombigree"?

Haplotype pedigree seems fine to me, but nowhere near as great as recombigree!

Like, you seriously think we should say "recombigree", or you think it's a great word?

I suppose we could write " In so doing we record the population pedigree as well as the outcome of each recombination; since this records genealogical history to the haplotype level we refer to this as the 'haplotype pedigree', or, whimsically, 'recombigree'. " but then not go on to actually use 'recombigree' a lot.

[ashander]

I think Kevin's comment was in a block

On Tue, Oct 24, 2017 at 2:06 PM, Peter Ralph notifications@github.com wrote:

I think to most people it would mean a pedigree annotated with disease status. How about "haplotype pedigree"? Or, "haplogree"? "recombigree"?

Haplotype pedigree seems fine to me, but nowhere near as great as recombigree!

Like, you seriously think we should say "recombigree", or you think it's a great word?

I suppose we could write " In so doing we record the population pedigree as well as the outcome of each recombination; since this records genealogical history to the haplotype level we refer to this as the 'haplotype pedigree', or, whimsically, 'recombigree'. " but then not go on to actually use 'recombigree' a lot.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/petrelharp/ftprime_ms/issues/8#issuecomment-339131310, or mute the thread https://github.com/notifications/unsubscribe-auth/AAfLONNg42z9YAuG_99jsl-HRaT_3IgAks5svlFzgaJpZM4QDXe3 .

[molpopgen]

I'm not sure it was. A chance to completely confuse the field with a fun new term may be too much to pass on.

On Tue, Oct 24, 2017 at 2:57 PM ashander notifications@github.com wrote:

I think Kevin's comment was in a block

[ashander]

OK, my 2c:

Haplotype pedigree is pretty clear and seems distinguishable from the ARG as the graph of haplotype relationships produced by an ARG. If we're to mention one of the whimsical ones, haplogree is better than recombigree as it's directly showing only relationships between haplotypes not recombinations (like a pedigree is showing individual relationships).

Other terminology ideas:

• ancestral inheritence graph

• haplotype ancestry

• Aside: the etymology of pedigree is not helping us here. It's from pied de gru or foot of a crane! https://blog.oup.com/2014/05/pedigree-etymology- word-origin/

On Tue, Oct 24, 2017 at 2:58 PM, Kevin R. Thornton <notifications@github.com

wrote:

I'm not sure it was. A chance to completely confuse the field with a fun new term may be too much to pass on.

On Tue, Oct 24, 2017 at 2:57 PM ashander notifications@github.com wrote:

I think Kevin's comment was in a block

On Tue, Oct 24, 2017 at 2:06 PM, Peter Ralph notifications@github.com wrote:

I think to most people it would mean a pedigree annotated with disease status. How about "haplotype pedigree"? Or, "haplogree"? "recombigree"?

Haplotype pedigree seems fine to me, but nowhere near as great as recombigree!

Like, you seriously think we should say "recombigree", or you think it's a great word?

I suppose we could write " In so doing we record the population pedigree as well as the outcome of each recombination; since this records genealogical history to the haplotype level we refer to this as the 'haplotype pedigree', or, whimsically, 'recombigree'. " but then not go on to actually use 'recombigree' a lot.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub < https://github.com/petrelharp/ftprime_ms/issues/8#issuecomment-339131310 , or mute the thread < https://github.com/notifications/unsubscribe- auth/AAfLONNg42z9YAuG_99jsl-HRaT_3IgAks5svlFzgaJpZM4QDXe3

.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/petrelharp/ftprime_ms/issues/ 8#issuecomment-339144386, or mute the thread https://github.com/notifications/unsubscribe- auth/AGHnH3NM_Xr24OtunN6cri-ZkxvVqkHHks5svlwNgaJpZM4QDXe3 .

--

Kevin Thornton

Associate Professor

Ecology and Evolutionary Biology

UC Irvine

http://www.molpopgen.org

http://github.com/ThorntonLab

http://github.com/molpopgen

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/petrelharp/ftprime_ms/issues/8#issuecomment-339146429, or mute the thread https://github.com/notifications/unsubscribe-auth/AAfLOO699WP1XmEf1vFc5ud6dfauvlAeks5svl2XgaJpZM4QDXe3 .

[petrelharp]

• Aside: the etymology of pedigree is not helping us here. It's from pied de gru or foot of a crane! https://blog.oup.com/2014/05/ pedigree-etymology- word-origin/

"nedigree", from "nest of a crane" (nid d' grue), which are made from lots and lots of (forking) branches?

[ashander]

Oh that's awesome!

On Tue, Oct 24, 2017 at 5:50 PM, Peter Ralph notifications@github.com wrote:

• Aside: the etymology of pedigree is not helping us here. It's from pied de gru or foot of a crane! https://blog.oup.com/2014/05/ pedigree-etymology- word-origin/

"nedigree", from "nest of a crane" (nid d' grue), which are made from lots and lots of (forking) branches?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/petrelharp/ftprime_ms/issues/8#issuecomment-339179493, or mute the thread https://github.com/notifications/unsubscribe-auth/AAfLONfomn6_jImAG_xWzcCjad_LWpL6ks5svoXBgaJpZM4QDXe3 .

[jeromekelleher]

Haplotype pedigree seems fine to me, but nowhere near as great as recombigree!

I was missing some <joke></joke> tags here as well. The whimsical words are good fun, but I don't think we should put them in the paper. It's all confusing enough as it is!

[petrelharp]

Haplotype pedigree seems fine to me, but nowhere near as great as recombigree!

I was missing some tags here as well. The whimsical words are good fun, but I don't think we should put them in the paper. It's all confusing enough as it is!

—

I agree; was hoping that brainstorming would come up with something more useful and less whimsical.

[petrelharp]

I'm going with "embelished pedigree" and/or "haplotype pedigree" on the current pass-through.

[jeromekelleher]

Sounds good to me.

[ashander]

over in #34 I edited @molpopgen 's use of 'ancestry tracking' to say 'pedigree recording'. I actually like the former a little better --- it is more descriptive to me of what we are actually doing in the simulation --- but the latter is more consistent with how we've described things elsewhere in the ms.

[molpopgen]

I tend to think of this as a "segment-wise ancestry tracking", leading me to write ancestry tracking each time.

[jeromekelleher]

I think we can use both 'ancestry tracking' and 'pedigree recording'; they're both descriptive of what we're doing.

[petrelharp]

Note for future selves: here is more discussion of the "ARG" issue: https://github.com/petrelharp/ftprime_ms/pull/6#discussion_r146101673