Handling Polymorphic States with Asymmetric Stepmatrix in TNT (Windows Interface)

[Serdar Sönmez]

Hello everyone,

I am currently facing a challenge in TNT regarding the treatment of polymorphic character states when using a custom asymmetric stepmatrix.

In my morphological dataset, I define a three-state character with the following interpretation:

0: Character is absent (i.e., the trait likely never evolved in this lineage)
1: Character is present (the only state I can directly observe)
2: Character was present but subsequently lost (inferred loss, based on Dollo logic)

In some taxa, the character is not observed, but it is unclear whether this is due to a true absence (0) or a secondary loss (2). In such cases, I intend to code the character as polymorphic (0&2) to explicitly represent this ambiguity while excluding the only observable state (1).

I would like to use a stepmatrix with the following cost scheme:

• 0 → 1 = 1

• 1 → 2 = 1

• 2 → 1 = 10 (to simulate Dollo irreversibility)

• 1 → 0 = 10

However, when using the Windows interface of TNT, polymorphic tip entries [0&2] seem to be automatically interpreted as uncertainties or expanded into additional states. A message appears during data entry indicating that further states are being added automatically. As a result, the asymmetric costs defined in the stepmatrix do not appear to be included during analysis.

It's possible that I'm misunderstanding the correct input format or missing a specific option to retain the polymorphic states and apply the intended cost structure. Therefore, I would greatly appreciate any advice on:

1. Whether TNT can maintain user-defined asymmetric stepmatrices when polymorphic character states are present at the tips.

2. Whether there is a recommended way to explicitly model ambiguous presence/absence without having them reduced to uncertainties.

3. Any workaround, scripting strategy, or undocumented functionality that may help preserve intended state transitions and costs in such cases.

Many thanks in advance for any clarification or suggestions.

Best regards,
Serdar SÖNMEZ, PhD
Professor, Adıyaman University

--

Prof. Dr. Serdar SÖNMEZ

Adıyaman Üniversitesi Fen Edebiyat Fakültesi Biyoloji Bölümü

Altınşehir Mh. 3005 Sokak No:13 02040 ADIYAMAN, TÜRKİYE

+90 416 223 38 00  / 4235