iiinova — where your digital self keeps evolving

James Fleming @jamesfvfleming.bsky.social · 22/09/2026

It had never really occurred to me that you can get full colour banknotes rather than shaded ones! I'm still in love with the Irish Punt coin designs before the transition to the Euro, especially the 1p with the Kells-Bird and the 50p Woodcock

010

James Fleming @jamesfvfleming.bsky.social · 16/09/2026

Pretty much! The work of a few incredibly talented taxonomists over the last 40 years especially has really accelerated our understanding of the group. Had to try and find this wonderful bit from Li et al 2025. Kinorhynchs Georg.

021

James Fleming @jamesfvfleming.bsky.social · 07/07/2026

I like to keep a histogram of how long it takes to find out whether I have an interview or not from certain jobs, just as a bit of fun. Helps keep that feeling of ever growing dread away, just a little bit.

A histogram of days from the application deadline until response from an employer in academia. The peak is left shifted (26-40 days following the deadline), but the extreme right of the graph contains a bin of 2 positions between 170 and 373 days.

010

James Fleming @jamesfvfleming.bsky.social · 03/07/2026

Finishing my #SMBE2026 trip to Copenhagen with a bit of a busman's holiday, but these are stunning.

040

James Fleming @jamesfvfleming.bsky.social · 08/06/2026

How robust *are* macrosyntenic analyses? We created a new method to test this. We created 100 random subsets of each analysis, and tested again. Did we always recover those relationships as significant? It turned out that we didn't! So despite looking nice, the original results weren't reliable!

A plot showing how reliable the inferences are. Despite "significant results" being recovered in the original analysis, and producing a diagram that looks great, we found that many of these relationships could not be recovered across our subsampling analyses, so can't be relied on.

121

James Fleming @jamesfvfleming.bsky.social · 08/06/2026

Macrosynteny is another method people are excited about lately. This is the idea that the order of genes on a chromosome could indicate shared evolutionary history, just like the evolution of a gene. In our case, the order of genes looked random. But the analysis said significant. So we look deeper!

An "oxford dot plot", showing where genes are matched to one another on different chromosomes between two animals.
The first graph shows a Lancelet (a vertebrate) and Mizuhopecten (a mollusc). There are dark areas and light areas, showing clear relationships.
The second and third panels show a rotifer vs. a chaetognath, with no clear relationships, despite being more closely related (in time)!

121

James Fleming @jamesfvfleming.bsky.social · 08/06/2026

What we found was that a lot of modern tools all pointed in one direction - Chaetognatha as sister to all the other Chaetognathifera, and Acanthocephala as sister to the seisonids (which in turn was sister to the bdelloids). The consensus between methods was actually really surprising!

A phylogenetic tree showing the results we recovered.

121

James Fleming @jamesfvfleming.bsky.social · 08/06/2026

Second... what is going on inside rotifers? There are three classes of rotifers - bdelloids, monogononts and seisonids. And Acanthocephala (sometimes, it is outside rotifers). But *how* these groups are related is unclear. So we wanted to test this...

A figure showing eight phylogenetic trees of Rotifera. As there are only 4 groups of Rotifera, this is basically every arrangement of the group!

131

James Fleming @jamesfvfleming.bsky.social · 08/06/2026

But while most everyone agrees all these groups are related *how* they are related is another matter! And this has big implications for how we understand their evolution! There are two major controversies here. First, did the Chaetognaths, or the Gnathostomulids diverge earlier?

A figure showing two phylogenetic trees of Chaetognathifera. One depicts Chaetognatha as the earliest diverging group, with Gnathostomulida as sister to Rotifera and Acanthocephala. Another, Gnathostomulida first and Chaetognatha as sister to Rotifera and Acanthocephala.

121

James Fleming @jamesfvfleming.bsky.social · 08/06/2026

For the round-up we'll start with the systematic, and end on the methodological! Chaetognathifera are a group of 4 (5?) phyla of invertebrates - Rotifers, Gnathostomulids, Micrognathostomulids, Chaetognaths and (sometimes) Acanthocephala (not pictured, talk about that later). They are very cute!

100

James Fleming @jamesfvfleming.bsky.social · 23/05/2026

So I'm already quoting one of my favourites down below, but this one is a brilliant work of art

010

James Fleming @jamesfvfleming.bsky.social · 19/05/2026

Crawled over the line to 1000 citations this week! Does it really mean anything? Not really. Does it make me feel a little better even though I don't have a permanent position yet? Absolutely!

020

James Fleming @jamesfvfleming.bsky.social · 15/05/2026

Not pictures, bag of scratchings. Living the dream

010

James Fleming @jamesfvfleming.bsky.social · 30/04/2026

But if we are comparing every strip head to head, we can't ignore its Nemesis

020

James Fleming @jamesfvfleming.bsky.social · 25/03/2026

I like that the Trams in Barcelona have Tram written on the side of them, just in case you forget what they are, or have never seen a Tram before.

A Barcelona metro tram, a light blue tram with "Tram" written on the top.

020

James Fleming @jamesfvfleming.bsky.social · 07/03/2026

This one is easy! Set around where I grew up, on the canals, I went to University and found out apparently nobody else watched this one?

010

James Fleming @jamesfvfleming.bsky.social · 26/02/2026

This is a really cute idea for the 30th Anniversary. But I can't just choose 1. All 3 of these deserve the top spot. www.serebii.net/pokemon30/#a...

020

James Fleming @jamesfvfleming.bsky.social · 01/02/2026

The reason that score varies is that the DE-Score normalises itself for the size of your dataset (the bottom bit), letting you compare saturation between datasets regardless of how their size. So it's simple! Input the alignment, look for Scores bigger than 0, then bigger than the Critical! 7/7

The DE-Score Equation again!
This time, we'll focus on the bottom of this division, which is 0.255n^-0.15. n is the number of sequences in the dataset, so this value was derived from our simulation data as a way to balance out the size of a dataset, to let you compare DE-Scores between datasets.

130

James Fleming @jamesfvfleming.bsky.social · 01/02/2026

When changes between categories are equally as likely as within categories (which is when the within/between ratio = 0.177), then the dataset is too noisy to be useful for phylogenetics! This is when the DE-Score equals 0. 5/7

An equation showing how the uninformative within/between category ratio was calculated.

The DE-Score equation! The top shows in brackets a pair-wise comparison of each amino acid in each position in a dataset, then subtracts 0.177 from that value, to give a distance from the uninformative ratio.

100

James Fleming @jamesfvfleming.bsky.social · 01/02/2026

We can divide amino acids into 6 categories based on their physicochemical properties - Dayhoff Categories. As changes in these amino acid categories can affect the structure and function of a protein, amino acid changes between categories should be rarer than changes within the categories. 4/7

A Figure showing the 6 dayhoff categories, and the amino acids that belong to them

100

James Fleming @jamesfvfleming.bsky.social · 31/01/2026

Sorry train fans. Boats are the superior form of public transportation

010

James Fleming @jamesfvfleming.bsky.social · 15/09/2025

Youtube outdoing itself with the thumbnail tonight.

A thumbnail of virtual singer Kasane Teto swearing in "Birdbrain"

010