Lastly, we show the usefulness of the model in stuyding empirical system of Anolis lizard and Viburnum plant diversification, and found evidence of local species diversity plays a major role in biogeographic radiation of these systems (10/10)
We use phylogenetic deep-learning for parameter estimation and model selection tasks. Using simulated data, we show that the method can recover most parameters accurately under the model, with diversity-dependent extinction being the hardest to estimate (9/10)
As an additional exploration, we also did some experiments on tree shape statistics generated under our model and found some interesting findings when the effect is weaker or stronger on a certain evolutionary process (8/10).
(2) only on within-region speciation, inbound dispersal, and extinction, (3) on all processes except between-region speciation. However, they all arrive at the same equilibrium dynamics (7/10)
For example: these three sets of parameters correspond to three different diversity-dependent dynamics. (1) diversity-dependence acts on all the evolutionary processes,
More importantly, each rate process has its own diversity-dependent parameter, meaning different combinations of these choices can impact the equilibrium diversity in a different way or (in fact) the same way (6/10).
This means, users do not need to estimate this parameter, which can sometime exhibit statistical bias. Rather, it can be derived directly from the rate parameters (5/10).
2) it is not assumed to be a parameter that influences the evolutionary rates directly, but rather an emergent property (consequence) of the data-generating process (rate parameters).
Some unique features of this parameter: 1) this is tied to each individual region, rather than as a global parameter,
In this work, we also define the concept of local equilibrium capacity under our model. This is the stationary phase of the process, in which the average number of local species remains unchanged over time under a given diversification model (4/10).
For example, we can include scenarios where high local species diversity in a specific region either has positive, negative or neutral feedback on shifts in extinction, within-region speciation, between-region speciation, or dispersal rates (3/10).
By that, the model is flexible in testing (inference) or simulating (data-generating) a range of evolutionary scenarios in which local species richness influences diversification dynamics through different biogeographic processes (2/10)
In its core, DDGeoSSE, is an event-based, fully generative phylogenetic diversification model that allows local species diversity to influence biogeographic rates of change.
Our model, DDGeoSSE, to the best of our knowledge, is the first to link species diversity, historical biogeography, and state-dependent diversification dynamics with the level of generality that SSE models can do (1/10).
Figure taken from Morlon et al. (2024)
www.annualreviews.org/content/jour...
New preprint on a new diversity-dependent biogeographic diversification model using deep learning. Joint work with my amazing advisor @landismj.bsky.social. Details of the paper in the comment section. #evobio
www.biorxiv.org/content/10.6...
Decided to add some non-academic content to my personal site. Right now, it features personal photos of (mostly Indonesian) food I made, batiks, and (anime) figures!
More here:
sites.google.com/view/albert-...
Some of my favorites Indonesian Batiks en.wikipedia.org/wiki/Batik_i...
Last January, I gave a seminar talk at Indonesia's National Research and Innovation Agency on mathematical and statistical approaches to study evolutionary processes.
The slides (in English) are available on my website for anyone interested: sites.google.com/view/albert-...
#evobio
Forests and mountains of West Timor
Beaches of West Timor - I
Orchids of West Timor, Indonesia - II
Orchids of West Timor, Indonesia - I
Cartoon of paleogeographic, biogeographic, and phylogenetic history of an island plant radiation, and plots of hypothesized relationships between regional features and biogeographic rates.
New preprint modeling biogeo diversification of Hawaiian Kadua 🌱🏝️🌋
w/ @ca-naturalist.bsky.social @sswiston.bsky.social @fabiology.bsky.social @phylogeny.bsky.social Warren Wagner, Bruce Baldwin, Ken Wood @ninaronsted.bsky.social @fzapata.bsky.social
biorxiv.org/content/10.64898/2025.12.16.694722
New preprint on a new infectious disease model with visitor dynamics using phylogenetic approach 😊
Giving a public lecture to math students and lecturers, introducing the use of mathematics in evolutionary biology and infectious disease modeling. It's good to contribute back to my almamater in Indonesia.
For Indonesian viewers, my materials can be accessed here: sites.google.com/view/albert-...
An honest mistake by mathematician 😆
Some screenshots from this year's Phylomania - great to see familiar faces and names, even through a virtual screen :)
Also, my talk can be accessed here: sites.google.com/view/albert-...
(preprint is hopefully out soon) #evobio
Phylomania will kick off in a few hours. Will be presenting my work with @landismj.bsky.social on a biogeographic model wiith diversity-dependence for modeling species diversification 😊
www.maths.utas.edu.au/phylomania/p...
(By the way, Joe Felsenstein is among the presenters)
#evobio
It always feels nice to be able to give something back to your alma mater. It has been 10 years since an 18 yo me completed his undergraduate and got introduced to the wonderful world of evolutionary biology and phylogenetic inference 😊
#evobio
Today at WashU’s EEB seminar, I presented a new species diversification model that incorporates regional species diversity into rates of evolution through time (a joint work w/ @landismj.bsky.social).
Slides here: sites.google.com/view/albert-...
Preprint hopefully will be here this month #evobio
