PhD Thesis Diversification Models and Neural Inference

This repository contains the complete LaTeX source, figures, and supporting files for my PhD thesis.

Thesis (front page)

Propositions

Publications

• Qin, T., Valente, L.†, & Etienne, R.† (2025). Impact of evolutionary relatedness on species diversification and tree shape. Journal of Theoretical Biology. https://doi.org/10.1016/j.jtbi.2024.111992
  🏆 Awarded the inaugural Denise Kirschner Best Student Paper Prize (2025).
  † Joint senior authors.

• Qin, T., van Benthem, K., Valente, L.†, & Etienne, R.† (2025). Parameter estimation from phylogenetic trees using neural networks and ensemble learning. Systematic Biology. https://doi.org/10.1093/sysbio/syaf060
  † Joint senior authors.

• Qin, T., van Benthem, K., Valente, L.†, & Etienne, R.† (2026). Identifying evolutionary relatedness effects on diversification from phylogenies using neural networks. bioRxiv. https://doi.org/10.64898/2026.01.12.699140
  † Joint senior authors.

Citation

@phdthesis{qin_diversification_2026,
  title  = {Diversification Models and Neural Inference},
  author = {Qin, Tianjian},
  school = {University of Groningen},
  year   = {2026},
}

Download the compiled PDF

Most recent build

This version is compiled automatically from the latest commit. The GHA workfow is built on top of latex-action.

• Read online: Click Here
• Direct download: Click Here
• Branch history: Click Here
• Build workflow: Click Here

Printed version

This version is consistent with the final printed PDF.

• PDF-for-print release: Not available yet

Packages and repos

evesim

The C++ optimized "eve" model is already published on CRAN.

• CRAN/evesim
• EvoLandEco/evesim

EvoNN

The illustrative implementation of the neural network parameter estimator is published in the EvoNN repository.

• EvoLandEco/EvoNN

Supplementary materials

Additional files and figures for the main thesis repository are maintained in the companion repository.

• EvoLandEco/Thesis-Appendix

Codebase

The codebase used throughout the thesis is deposited in the eveGNN repository.

• EvoLandEco/eveGNN

Use on Overleaf

Setup note: This project is intended to be compiled on Overleaf and with GitHub Actions.

Option A: Upload a ZIP

1. Click Code → Download ZIP on GitHub.
2. In Overleaf, create a new project and Upload Project (ZIP).
3. Set the Main file to dissertation.tex.

Option B: GitHub–Overleaf sync

If you use Overleaf’s GitHub integration, link this repository to your Overleaf project and keep the repo as the single source of truth.

Repository structure

• dissertation.tex — main entry file
• acks/ — acknowledgements sources
• conclusion/ — conclusion sources
• summary/ — summary sources
• title/ — title page sources (and cover preview image)
• cv/ — CV sources
• publications/ — publication sources
• introduction/ — introduction sources
• chapter*/ — chapter sources
• dissertation.cls — custom class/style
• dissertation.bib — references

License

• Code and build/config files: see LICENSE

Acknowledgements

This thesis repository started from the LaTeX template by Moritz Beller

I am grateful for this solid foundation. Since adopting it, I have modified and extended the codebase (structure, styling, build/CI workflow, and compatibility fixes). Any remaining issues are my own.