Corrections

Below is a running list of minor corrections for our papers (paper numbers correspond to those on the publications page). If you identify any other issues, please send me an email!

[2]

Squirrel: Reconstructing Semi-directed Phylogenetic Level-1 Networks from Four-Leaved Networks or Sequence Alignments

Molecular Biology and Evolution, 2025

• Materials and Methods, \(\delta\)-heuristic, Steps I–III: the interpretation of the \(\delta\)-value is reversed in the text. A tree-like quartet has \(\delta\) close to 0, so it should read as follows. In Step I, if \(\delta_{\{a,b,c,d\}}\) is close to 0, we expect the split \(ad|bc\) (the text says "close to 1" and "\(ab|cd\)"). In Step II, \(S_\lambda\) contains the 4-taxa subsets whose \(\delta\)-value is at most \(\lambda\), and \(F_\lambda\) those whose \(\delta\)-value is larger than \(\lambda\) (the text says "at least" and "smaller"). In Step III, a 4-taxa set in \(S_\lambda\) is assigned the split \(s \in \{ab|cd,\, ac|bd,\, ad|bc\}\) for which \(h_s\) is smallest (the text says "the highest").
• Supplementary Material, proof of Lemma A.3: the words "opposite" and "neighbours" are swapped. For \(i = 1\), \(y_i\) and \(y_j\) are neighbours in \(\binom{s_{1j}}{2} + \binom{t_j}{2}\) of the 4-cycles and opposite in \(s_{1j} \cdot t_j\) of them; for \(i \geq 2\), they are neighbours in \(r_i + t_j\) of the 4-cycles and opposite in \(s_{ij}\) of them. The displayed formulas are correct.
• Supplementary Material, proof of Lemma A.5, Case 1, first condition: the expression simplifies to \((k-j)\cdot(2n - k - j + 1)\), with a minus sign before \(k\); the text says "\((k-j)\cdot(2n + k - j + 1)\)". The inequality still holds, since \(k + j \leq 2n - 3\).