Mutation of Skew-Symmetrizable Matrices

• Mutation of skew-symmetrizable matrices is an involutive operation generalizing Fomin–Zelevinsky mutations to matrices with a symmetrizer in cluster algebra theory.
• The mutation rule uses a precise formula to adjust matrix entries and preserves key invariants, such as the Markov constant in rank-3 cases.
• Applications span weighted quivers, species with potentials, and unfoldings, linking algebraic, combinatorial, and geometric methods.

A mutation of skew symmetrizable matrices is an involutive operation, originally introduced in the context of cluster algebras, that generalizes Fomin–Zelevinsky mutations for skew-symmetric matrices to the broader class of skew-symmetrizable matrices. These matrices arise naturally in a variety of algebraic, combinatorial, and geometric settings, notably in the theory of cluster algebras, representation theory, and Lie theory. The theory encompasses explicit mutation formulas, invariants, classification of mutation classes, connections with diagrams and weighted quivers, as well as categorifications using quivers with potentials, species, and geometric realizations.

1. Skew-Symmetrizable Matrices and Mutation Rule

A real or integer $n \times n$ matrix $B=(b_{ij})$ is skew-symmetrizable if there exists a diagonal matrix $D=\mathrm{diag}(d_1,\dots,d_n)$ with $d_i>0$ (real or integer) such that $D B$ is skew-symmetric: $(D B)_{ij} = - (D B)_{ji} \quad \forall i, j.$ Equivalently, $d_i b_{ij} = - d_j b_{ji}$. The matrix $D$ is called a symmetrizer and is typically taken to have minimal positive integer entries in the cluster algebra context.

Given such a $B$, for each index $k \in \{1, \dots, n\}$, the mutation $\mu_k(B)$ is an $n \times n$ matrix $B'=(b'_{ij})$ defined by

$b'_{ij} = \begin{cases} - b_{ij}, & \text{if } i = k \text{ or } j = k,\[6pt] b_{ij} + \dfrac{ |b_{ik}| b_{kj} + b_{ik} |b_{kj}| }{2}, & \text{otherwise}. \end{cases}$

This rule generalizes the Fomin–Zelevinsky mutation formula for skew-symmetric matrices. One readily checks that $\mu_k$ is involutive ($\mu_k^2 = \mathrm{id}$) and preserves skew-symmetrizability with the same symmetrizer $D$ (Akagi, 2024, Felikson et al., 2010, Seven, 2010).

2. Weighted Quivers, Diagrams, and Their Mutation

Each skew-symmetrizable $B$ corresponds to a weighted quiver or diagram $Q(B)$ whose vertices index the rows and columns of $B$. For each pair $i \ne j$, if $b_{ij} > 0$, the quiver has an arrow $i \to j$ with the ordered pair of weights $(|b_{ij}|, |b_{ji}|)$, often encoded as a single edge label $w_{ij}=|b_{ij}b_{ji}|$.

Mutation of $B$ induces a mutation of the quiver in which:

• All arrows incident to $k$ are reversed,
• For each pair $i \to k \to j$, the number of arrows from $i$ to $j$ is adjusted according to the combinatorial rule corresponding to the matrix mutation formula,
• (Optionally) Maximal collections of 2-cycles are removed in quiver-theoretic settings.

This combinatorial operation underpins the mutation class structure and is tightly linked to the block decomposition and unfolding theory of cluster algebras (Felikson et al., 2010, Seven, 2010, Seven, 2012).

3. Mutation Classes and Classification: Rank-3 and Beyond

For $n=3$, every skew-symmetrizable matrix $B$ can be classified up to mutation equivalence using combinatorial, algebraic, and numerical invariants:

• Cyclic vs. acyclic diagrams: A cyclic diagram has all six off-diagonal $b_{ij}$ sharing the same sign pattern, causing the associated quiver to form a directed cycle; otherwise, it is acyclic.
• Minimal representative: Each mutation class has a unique (up to equivalence) minimal representative characterized by the radical weights of the edges, with explicit inequalities discriminating cyclic and acyclic cases (Seven, 2010).
• Markov constant: For rank-3 cyclic $B$, an explicit invariant—the Markov constant

$C(B) = x x' + y y' + z z' - |x y z|,$

in appropriately parameterized form—is preserved under all mutations and determines whether the mutation class remains cyclic under every sequence (cluster-cyclic) (Akagi, 2024). The precise classification is given by $C(B) \leq 4$ together with $x x', y y', z z' \geq 4$, or, using symmetric double-sided coordinates $(p, q, r)$, $p, q, r \geq 2$ and $p^2 + q^2 + r^2 - p q r \leq 4$.

In higher rank, the finiteness classification for mutation classes is based on the existence of $s$-decompositions or exceptional diagrams, as in the theorem:

• $B$ is of finite mutation type if and only if its diagram is $s$-decomposable or mutation-equivalent to one of the known exceptional types (notably $G_2$, $F_4$ and their variants) (Felikson et al., 2010, Gu, 2012).

4. Unfoldings, Species, and Categorification

Unfolding refers to the process of embedding a skew-symmetrizable matrix $B$ into a higher rank skew-symmetric matrix $C$ such that the mutation dynamics of $B$ correspond to a symmetric folding of the mutations in $C$. For $s$-decomposable diagrams, such unfoldings always exist and can be constructed explicitly using block decomposition (Felikson et al., 2010, Gu, 2012).

Species with potentials generalize quivers with potentials to the skew-symmetrizable context:

• Species assign, for each $i$, a division ring (or group algebra), and for each edge, a bimodule structure. Potentials are defined as formal linear combinations of cyclic paths in the path algebra of the species.
• Mutation at a vertex involves a combinatorial modification of the species structure and mutation of the potential, together with a reduction step to eliminate 2-cycles.
• The translation from species with potential back to cluster algebraic data yields exchange matrices, $F$-polynomials, and $g$-vectors corresponding to the combinatorics of mutations (Demonet, 2010, Labardini-Fragoso et al., 2013, López-Aguayo, 2017).
• The existence of non-degenerate potentials is now established for large classes of skew-symmetrizable matrices, including those not globally unfoldable or not strongly primitive, when divisibility conditions on the symmetrizer and entries are met (López-Aguayo, 2017).

5. Mutation Invariants and Reflection Group Perspectives

Mutation classes admit a variety of combinatorial and algebraic invariants:

• Besides the Markov constant in rank 3, there exist $n$-vertex diagrams invariants such as the sorted tuple of greatest common divisors of incident weights per vertex, which remain unchanged under all mutations (Seven, 2010).
• More generally, cluster algebras with skew-symmetrizable exchange matrices are linked to Kac–Moody root systems and reflection groups. Every acyclic $B$ yields a generalized Cartan matrix $A$, with the Weyl group $W(A)$ given a presentation via relations derived from mutation classes: orders of products of generating reflections correspond to diagram weights, together with additional relations for oriented cycles (Seven, 2012).
• Oriented cycle relations provide strong constraints and, for acyclic diagrams with all edge weights $\geq 4$, mutation cannot produce smaller weights (Seven, 2012).

6. Broader Implications in Cluster Algebras and Geometry

The mutation theory for skew-symmetrizable matrices plays a central role in the structural theory of cluster algebras:

• Classification of mutation-finite types corresponds to triangulated marked bordered surfaces and their topological types (Felikson et al., 2010).
• Structural results about maximal green sequences and reddening sequences depend on the combinatorics of mutation classes, with rank 3 providing a foundational laboratory. For example, no maximal green sequence exists in mutation-cyclic rank-3 classes (Seven, 2012), and in Lie-theoretic settings, all string-diagram exchange matrices admit reddening sequences (Cao, 2022).
• For real-weighted matrices, finite mutation type is characterized and connected to the geometry of acute-angled simplexes in reflection groups, and the exchange matrices often admit geometric realizations via Gram matrices in vector spaces of determined signature (Felikson et al., 2019).
• The construction extends to matrices over group rings and incorporates quivers with symmetries and their folding, with generalizations of the mutation rule now available in this noncommutative framework (Kaufman et al., 22 Jan 2026).

7. Examples, Computations, and Practical Methods

Explicit examples and algorithms are available for understanding and calculating mutations:

• Step-by-step examples of mutation for rank-3 cyclic matrices, including the appearance and persistence of the Markov constant, demonstrate practical computation (Akagi, 2024).
• The decomposition (unfolding) algorithm for $s$-decomposability operates in linear time, verifying whether a given skew-symmetrizable matrix admits a block-decomposable unfolding associated to a marked surface (Gu, 2012).
• For matrices admitting a string-diagram realization, one obtains a uniform reason for the existence of reddening sequences and canonical potentials, encompassing many important classes arising in Lie theory (Cao, 2022).
• Worked examples in the group-ring setting illustrate the necessity of generalized sequences and the appearance of novel phenomena in the presence of nonzero diagonals and higher symmetry (Kaufman et al., 22 Jan 2026).

---

Table: Key Mutation Formulas for Skew-Symmetrizable Matrices

Context	Mutation Formula (for $b'_{ij}$)	Reference
Standard (real/integer entries)	$$\displaystyle b'_{ij} = \begin{cases} -b_{ij}, & i = k \text{ or } j = k,\ b_{ij} + \frac{|b_{ik}| b_{kj} + b_{ik} |b_{kj}|}{2}, & \text{otherwise} \end{cases}$$	(Felikson et al., 2010, Akagi, 2024)
Rank-3 Markov constant	$C(B) = x x' + y y' + z z' - |xyz|$, invariant under mutation	(Akagi, 2024)
Group ring entries	$b'_{ij} = -b_{ij}$ if $i=k$ or $j=k$; otherwise $$b_{ij} + [b_{ik}]_+ [b_{kj}]_+ - [b_{ik}]_- [b_{kj}]_-$$	(Kaufman et al., 22 Jan 2026)

---

The mutation theory of skew-symmetrizable matrices thus provides a unifying generalization of quiver mutation, supporting a variety of representations, categorifications, and geometric models. It is fundamental for the analysis, classification, and construction of cluster algebras and their many applications in algebra, combinatorics, topology, and mathematical physics.