Trace all changes on trees

[Christopher Griffin]

Hi all,

Very excited for the new version of Mesquite! I'm sure I will use it as much as I have the previous version. 

I was wondering if there was any chance of the visual 'trace all changes' function from MacClade—wherein each character transition was visible on each tree—coming back for Mesquite v.4. This is an extremely useful function for me (see here and here for examples of how this has been used in analyses), to the point that I bought an old MacBook off Facebook Marketplace just to run MacClade so that I could do this! 

I understand there is a table 'trace all changes' function, but having it visually plotted on the tree would be very beneficial. I know of several colleagues who have been interested in performing the analyses in the linked papers, but have been put off by this feature not being easily accessible anymore.  

Thanks,

Chris

[Wayne Maddison]

Chris, thanks for nudge. This has been requested several times over the years, but we've been reluctant largely because of the messiness with ambiguity in the parsimony reconstruction, but also because the graphical flexibility of Mesquite's tree drawing meant that it would be very hard to execute in general. Back in the pure parsimony days, and with MacClade's more constrained drawing system, it was worth the pain to deal with it. But times have changed, and Mesquite has so many other things going on that a graphic Trace All Changes never made it to the top of the priority list. However, we've already made a choice that eases the graphics problem, so we can think about it for 4.x (maybe 4.1).

How would you feel if it showed only unambiguous changes, i.e. it would show nothing if the position of the change is equivocal?

[Christopher Griffin]

Hi,

I think even unambiguous changes would be extremely useful! Thanks for considering 

[Kipling W Will]

I have to agree with Chris on this. My workaround is to export to WinClada and view character state optimizations there. It only does unambiguous, acctran, or deltran, but it is quite helpful.

Kip

--
You received this message because you are subscribed to the Google Groups "Mesquite Project" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mesquite-proje...@googlegroups.com.
To view this discussion, visit https://groups.google.com/d/msgid/mesquite-project/4fd7c691-8877-4db1-aa20-37fc5541ccbdn%40googlegroups.com.

[Wayne Maddison]

Chris and Kip, you've convinced us to make some progress on this. Coming soon:

This will, at least in 4.0, most likely be restricted to parsimony-unambiguous changes.  How do you feel about that?

We are hoping to release a beta2 very soon, and this will be included. We'd very much appreciate your feedback about what features you might like to see added. We can't promise too many changes, but if it's something simple we might be able to.  -- Wayne

[Kipling W Will]

I think it will be great. Thanks for the effort.

Kip

--
You received this message because you are subscribed to the Google Groups "Mesquite Project" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mesquite-proje...@googlegroups.com.

To view this discussion, visit https://groups.google.com/d/msgid/mesquite-project/bf79f415-5f65-459e-b639-015c33e09d3fn%40googlegroups.com.

[David Maddison]

One thing to note is that this is truly unambiguous changes - not the not-really-unambiguous changes chosen by the arbitrary acctran/deltran options.  So expect fewer changes being shown than what you are used to if you use acctran/deltran options elsewhere. 

We'll try to add some other options in the future, maybe eventually getting to the full complexity displayed by MacClade.  

To view this discussion, visit https://groups.google.com/d/msgid/mesquite-project/CAN4RUMhdD9Vi2qcD-zEFPKQDPtWhtMCZezM-b2tKOtMUs8SPRQ%40mail.gmail.com.

[Christopher T. Griffin]

This looks like it will be very useful!

 

-- 

Chris Griffin

Assistant Professor of Geosciences

Princeton University

www.griffinlabpaleo.com

 

 

From: mesquite...@googlegroups.com <mesquite...@googlegroups.com> on behalf of David Maddison <bemb...@gmail.com>
Date: Saturday, June 7, 2025 at 12:13 PM
To: Kipling W Will <kip...@berkeley.edu>
Cc: Wayne Maddison <wayne.m...@ubc.ca>, Mesquite Project <mesquite...@googlegroups.com>
Subject: Re: Trace all changes on trees

One thing to note is that this is truly unambiguous changes - not the not-really-unambiguous changes chosen by the arbitrary acctran/deltran options.  So expect fewer changes being shown than what you are used to if you use acctran/deltran options elsewhere. 

 

We'll try to add some other options in the future, maybe eventually getting to the full complexity displayed by MacClade.  

 

 

 

On Sat, Jun 7, 2025 at 7:29 AM 'Kipling W Will' via Mesquite Project <mesquite...@googlegroups.com> wrote:

I think it will be great. Thanks for the effort.

Kip

 

On Fri, Jun 6, 2025 at 4:44 PM Wayne Maddison <wayne.m...@ubc.ca> wrote:

Chris and Kip, you've convinced us to make some progress on this. Coming soon:

 

This will, at least in 4.0, most likely be restricted to parsimony-unambiguous changes.  How do you feel about that?

 

We are hoping to release a beta2 very soon, and this will be included. We'd very much appreciate your feedback about what features you might like to see added. We can't promise too many changes, but if it's something simple we might be able to.  -- Wayne

 

--
You received this message because you are subscribed to the Google Groups "Mesquite Project" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mesquite-proje...@googlegroups.com.
To view this discussion, visit https://groups.google.com/d/msgid/mesquite-project/bf79f415-5f65-459e-b639-015c33e09d3fn%40googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Mesquite Project" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mesquite-proje...@googlegroups.com.
To view this discussion, visit https://groups.google.com/d/msgid/mesquite-project/CAN4RUMhdD9Vi2qcD-zEFPKQDPtWhtMCZezM-b2tKOtMUs8SPRQ%40mail.gmail.com.

--
You received this message because you are subscribed to a topic in the Google Groups "Mesquite Project" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/mesquite-project/POH4NYMHcCQ/unsubscribe.
To unsubscribe from this group and all its topics, send an email to mesquite-proje...@googlegroups.com.
To view this discussion, visit https://groups.google.com/d/msgid/mesquite-project/CAKL%3DO0MoJF-XbuG6%2B1F2j155eDqq4Y9mZ-6UKEjkFLeq2g4k2w%40mail.gmail.com.

[Wayne Maddison]

Beta 2 released with new Trace All Changes! In fact, its name has changed from Trace All Characters, because the latter didn't try to talk about changes. This is a first draft, and we expect it to change with your feedback.

• Available in Analysis:Tree menu in the Tree window. 
• Options controlled in Trace_All menu
• Two views, the display on the tree (the new part), and the table window (the old Trace All Characters).
• Options for tree display:
  • Unambiguous changes only
  • All changes permitted by parsimony. This marks whenever parsimony permits a change on the branch. This avoids the arbitrariness of ACTRAN and DELTRAN, but it means that one can't interpret the set of bars for a character as illustrating a single scenario of evolution. When there are alternative placements of apomorphies, ALL will be shown at once.
• To view bars for only one or a set of characters, select them (e.g. in the matrix or List of Characters window) and choose the menu item Selected Characters Only.
• Click on the bars to see details. A drop down menu will appear
• To trace a single character, select the menu item that drops down after you've touched on a bar

One interface change is planned but not yet implemented: to show, instead of boxes, just ticks without the colours.

Please give your feedback about changes you'd like to see. We're not likely to program ACCTRAN and DELTRAN soon (if at all), but we'd be happy to make some small interface changes.

Wayne