Tandem clustering of toxin genes exceeds gene-family-size expectations in four venomous bilaterians

Toxin genes of venomous animals often lie in tandem arrays, which are commonly read as a signature of venom-specific evolution. However, large gene families of any function also cluster, so how much of this architecture is specific to toxins remains unclear. In this study, we compared toxin clustering in five venomous genomes with random genes and with size-matched non-toxin families. It is anchored on a new chromosome-level genome of the centipede Scolopendra mutilans, whose venom-gland proteome (12 individuals) detected 163 of 188 curated toxins and resolved 141 to single loci. Against random genes, the odds of a toxin gene having another toxin gene within 100 kb were 14-31 times higher in four bilaterians (centipede, snake, spider and scorpion). Against size-matched non-toxin families, the odds ratio fell to about 2-6 but remained above 1 under three gene-family definitions and, under the primary definition, for windows of 25-200 kb, after matching on within-family sequence similarity and, in the snake, with published venom-gene annotations. In the sea anemone, evidence for an excess was insufficient. The excess was spread across many toxin families and was carried mainly by adjacent homologous copies. Duplication topology and expression did not distinguish toxin from control tandem families, and in a congener toxin arrays were recovered less often and copy-number comparisons were uncertain. Thus, a random-gene baseline overstates the exceptionality of toxin clustering, but toxin families keep more homologous copies in local arrays than equally sized families; whether such arrays form more often or persist longer remains open.