iiinova — where your digital self keeps evolving

And huge shoutout to @lagiven.bsky.social! She’s contributed a ton to my various dissertation projects, and it was a special to see her lead a project of her own. She’s planning to apply to PhD programs this fall!

020

As always, big thanks to @mollyschumer.bsky.social for allowing us freedom to see this project from conceptualization to draft and being extremely helpful whenever we needed, especially in the later stages. Lyle and I both learned a lot throughout the process!

More generally, we show how long-read sequencing can reveal substantial variation linked to sex chromosomes and sex-linked traits including inversions, ampliconic genes, and satellite DNA repeats.

So, is xmrk the patterning gene? We show that both patterns are linked to distinct haplotypes on the X and Y containing xmrk, a novel finding in this species, but we argue that more data are needed to settle this longstanding debate.

Closer inspection revealed these Rex and Helitron elements were truncated, fused, and tandemly repeated to form a novel class of composite satellites. xmrk was flanked on either side by large satellite blocks. Higher order satellite repeat structures varied substantially between chromosomes.

250kb inset showing self-alignment of “block-like” satellite regions within ampliconic array near xmrk. Top plot is of self-alignments with >100 bp of perfect identity. Highly self-identical regions consist of elevated repeat DNA content, which form tandemly arrayed composite satellite sequences. Shown below the inset is a zoomed-in view of 5 satellite monomers, consisting of truncated Rex3 elements, a complex satellite chimeric unit, a Helitron4-satellite chimera, and several unknown elements

130

Analysis of Y chromosomes revealed high levels of structural variation. In addition to the ampliconic regions containing key genes, we identified some “block-like” elements enriched with unusual repeats like Rex elements and Helitrons, a finding enabled by a new pan-Xiphophorus repeat library.

Composite plots showing distinct structural haplotypes in 4 Mb on the sex chromosome. Top panel shows StainedGlass self-alignments, colored by percent-identity. The first two tracks show sequence class annotations and intervals defined by GWAS. The ampliconic region containings xmrk (red dots) as well as mc4r (purple dots) Repeat content is elevated in this region.

We then explored variation on the sex chromosome near xmrk and more broadly. The X and Y chromosomes were largely co-linear, but the Y had a ~1mb inversion relative to the X. xmrk falls next to this inversion in an ampliconic region that does not reliably align between haplotypes.

Alignments between chromosome 21 haplotypes 1 and 2 are largely syntenic outside of the region associated with sex, which shows elevated rearrangement including an inversion (green) between X and Y chromosomes

To identify potential functional changes associated with spotting phenotype, we first looked at xmrk coding variation. xmrk proteins from the different spotted phenotypes differ at 3 amino acids, all within or near the predicted trans-membrane domain.

Non-synonymous amino acid substitutions plotted on the xmrk tree (left). Marmoratus has two private substitutions and spotted side has one. The derived substitutions associated with spotted side (blue) and marmoratus (red) are in or near the predicted trans-membrane domain of the protein (right).

To explore variation between haplotypes, we conducted PacBio HiFi sequencing and assembled another 8 sex chromosomes. Lyle found lots of copy number variation near the GWAS peak, but the only clear pattern was that xmrk was present in each spotted individual and absent in the unspotted fish.

Copy number variation of predicted protein-coding genes and lnc-RNAs within 500 kb of the GWAS peak for each haplotype, listed in alphabetical order. Numbers denote gene copy number and colors denote phenotype (blue—spotted side; red—marmoratus; grey—unspotted).

Lyle then performed GWASes for pigment patterns and also for sex. We found that the three signals co-localized, and the pigmentation traits mapped near xmrk. In total, she did 87 GWASes(!) using different reference genomes to investigate short-read mismapping biases.

Case-control GWAS identify a region on chromosome 21 associated with spotted side (top), marmoratus (middle), and sex (bottom). Left panel shows the whole genome, and middle panel shows a 4 Mb region surrounding the peak signal on chromosome 21 with annotations for protein-coding genes shown above (xmrk in red).

We next wanted to identify the genetic basis. But sex chromosomes are hard and xmrk is unassembled in most other species. I used the lab’s new P2 sequencer to run ONT ultralong sequencing. The resulting assembly was gapless and T2T except for 2 rDNA arrays. This fish had an X-linked xmrk (red)!

A gapless, near complete genome from a female with spotted side. Chromosomes are shaded by density of coding sequence in 50 kb windows (white-low, black-high). Caps on the end of chromosomes represent telomeres. 22 of 24 expected chromosomes are telomere-to-telomere (T2T) while chromosomes 18 and 23 end instead in an rDNA array (shown in blue). The primary assembly of chr-21 contains two copies of an xmrk-like gene, representing a putative xmrk (shown in red; position 22.9 Mb) and its paralog egfrb (shown in orange; position 24.9 Mb).

Despite occurring on different locations of the flank, Lyle showed the cells making up both spotting patterns are macromelanophores with similar diameters, and these are different than the smaller cells making up the normal background pigmentation.

Scales from individuals with spotted side, marmoratus, and no spots show macromelanophores and micromelanophores. Comparisons of cell diameter indicate macromelanophores are significantly larger than micromelanophores but do not significantly differ between phenotypes

To ask if these patterns were distinct, Lyle painstakingly quantified positions of spots on each fish… she referred to this process as “coloring time”. She performed an analysis on these pattern masks showing the phenotypes separate in pattern PC space.

Heatmaps of spots from individuals with spotted side (top) and marmoratus (bottom). A principal component analysis shows two distinct clusters, corresponding to spotted side and marmoratus, that separate along PC1.

130

Here, we investigate the genetic basis of macromelanophore patterns in the mountain swordtail X. nezahualcoyotl. The idea for this project started in the field, when we saw 2 seemingly distinct melanic patterns segregating in a new population. One of these phenotypes appeared male-limited.

Two classes of polymorphic spotting patterns segregate in males and females, matching descriptions of the patterns “spotted side” and “marmoratus.” Spotted side and marmoratus can co-occur in males, but marmoratus appears to be absent in females. Illustrations by Dorian Noel

For ~30 years, a debate has centered around the gene xmrk, whose misregulation causes melanoma in hybrids. There are two opposing hypotheses: 1) xmrk presence patterns macromelanophores vs 2) xmrk is tightly linked to an unidentified patterning gene. There is compelling evidence for both models.

In the genus Xiphophorus, variation in sex-linked melanic patterns have attracted attention for nearly a century (e.g., Gordon 1927 Genetics). These patterns are made up of large pigment cells called “macromelanophores”. But the genetic basis of macromelanophore pattern variation remains unclear.

Illustration of 2 macromelanophore patterns in Xiphophorus maculatus, described by Myron Gordon in 1927

130

Sex-linked traits are widespread, but sex chromosomes and the traits they encode are difficult to study. Long-read approaches allow us to “see” the full spectrum of genetic variation, providing new opportunities to advance our understanding of sex-linked traits.

Joana L. Rocha @joanocha.bsky.social · 08/06/2026

Reposted by Tris Dodge

We are excited to share our pan-pangenome paper! This is a long time coming work of my postdoc with the incredible @psudmant.bsky.social, and a stellar group of people! I hope you enjoy reading it! And I am excited to talk more more about it this week at #PEQG26!! www.biorxiv.org/content/10.6...

biorxiv.org

A Pan-pangenome illuminates complex structural variation and selection in humans, chimpanzees, and bonobos

Complete, haplotype-resolved genome assemblies have provided unprecedented insight into the evolution of structurally complex, rapidly evolving regions of human genomes; however, population-scale pang...

15825

Tris Dodge @tododge.bsky.social · 19/05/2026

We see similar off-target associations/mapping bias in fish! For us, if we do GWAS to one haplotype, we get a million hits, and if we map to the other hap, we get one. This such an important and under-appreciated problem. Congrats on the preprint and proposing a solution!

John Lovell @jotlovell.bsky.social · 19/05/2026

Reposted by Tris Dodge

We are often asked whether pangenomes are worth the hassle. The files can be large and methods to explore them are much more challenging than linear references. Earlier this week, we preprinted a clear example of how and why pan-genome graphs are so important for trait discovery.

13423

Tris Dodge @tododge.bsky.social · 19/05/2026

Thrilled to present my dissertation research at the Hamilton Symposium! We used approaches including GWAS, long-read sequencing, phylogenetic methods, and behavior experiments to uncover the genetic basis and evolutionary history of a shared polymorphism. Can’t wait to share what we’ve learned!

Picture of three diverged Xiphophorus species that display the false gravid spot, a sexual mimicry polymorphism that allows males to mimic females.

0226

Reposted by Tris Dodge

Kelsie Hunnicutt @kelsiehunnicutt.bsky.social · 12/05/2026

Thrilled that our new review article on animal hybridization is now up on EcoEvoRxiv! 🎉 Working on this review with @mollyschumer.bsky.social was a blast, and it was such a wonderful opportunity to catch up on all the amazing work in animal hybridization over the last few years!

A world map showing the distribution of species studied in hybridization research

04619

Tris Dodge @tododge.bsky.social · 22/04/2026

But the local adaptation and balancing selection communities do not seem to talk as much as I would expect. I wonder what this would look like if quantified, similar to what others have shown for balancing selection theoreticians and empiricists (onlinelibrary.wiley.com/doi/10.1111/...)

onlinelibrary.wiley.com

A century of theories of balancing selection

Traits that affect organismal fitness are often highly genetically variable. This genetic variation is vital for populations to adapt to their environments, but it is also surprising given that natur....

020

Tris Dodge @tododge.bsky.social · 22/04/2026

I see no real distinction between local adaptation and balancing selection, as long as the magnitude of gene flow is sufficient to approximate a panmictic population