This repository contains the original software (since 2017) associated with the publication:
Evolutionary model of protein secondary structure capable of revealing new biological relationships (publicly introduced in ISMB 2017, publicly released in February 2019 and formally published in May 2020)
This study introduced the first 3D structural evolutionary model that treats DSSP-defined protein secondary-structure states as evolving traits, building a framework for analysing 3D structural variation and uncovering biological relationships. Crucially, it provided an early computational basis for subsequent findings, revealing—for the first time—a functional link between plant Toll/interleukin-1 receptor (TIR) domains and enzymatic activity, which was later experimentally and structurally validated.
The secondary structure state (DSSP) evolutionary model was first introduced at the 2017 ISMB conference in the poster:
“Inferring protein phylogeny by modelling the evolution of secondary structure”
The enzymatic activity function was subsequently validated and published in "NAD+ cleavage activity by animal and plant TIR domains in cell death pathways, Science".
This MATLAB-based software package enables:
- Construction of phylogenetic trees based on secondary structure (DSSP)
- Estimation of evolutionary distances between proteins based on secondary structure (DSSP)
- Prediction of ancestral secondary structure states based on secondary structure (DSSP)
The toolkit is designed to extract evolutionary signals embedded in protein secondary structures—signals that are often overlooked by traditional sequence-based methods. It provides an alternative framework for protein phylogenetic analysis by modelling the evolution of structural features.
If you use this software, please cite the following publication:
Lai, J. S., Rost, B., Kobe, B., & Bodén, M. (2020).
Evolutionary model of protein secondary structure capable of revealing new biological relationships.
Proteins: Structure, Function, and Bioinformatics, 88(9), 1251–1259.
https://doi.org/10.1002/prot.25898
A publicly available preprint of the Lai et al. manuscript is accessible via bioRxiv:
Further enzymatic functions of TIR domains have since been experimentally validated.
Horsefield, S., Burdett, H., Zhang, X., et al. (2019).
NAD+ cleavage activity by animal and plant TIR domains in cell death pathways.
Science, 365(6455), 793–799.
https://doi.org/10.1126/science.aax1911
The full dataset associated with the manuscript is publicly available and can be accessed at:
Manuscript Dataset (140MB)
Feel free to submit pull requests for improvements.
This work established a protein evolutionary transition probability model centred on DSSP secondary-structure states, using carefully curated high-resolution protein crystal structures for model construction and validation. The methodological framework followed the core logic of classical evolutionary models such as Dayhoff’s PAM and JTT. Because DSSP states are strongly related to φ/ψ torsion-angle space, the model captures evolutionary transition patterns between secondary-structure states while indirectly reflecting associated changes in local backbone conformation. Related results were presented as a poster at ISMB 2017 (3DSIG).
Because the number of available high-resolution crystal structures was insufficient to comprehensively explore all joint DSSP–amino acid state transitions, I extended the analysis to a combined structural–sequence representation. At ISMB 2019, I presented a 60-state evolutionary substitution matrix for protein alignment, defined by three predicted secondary-structure states combined with 20 amino acid states. This work was later incorporated into my PhD thesis, where I further showed that a joint structural-property × amino-acid index can improve protein alignment performance by capturing coupled evolutionary patterns that are not represented by sequence information alone.
60-state evolutionary substitution matrix alignment and its early NMR-supported SARM1/ARM binding-site context
My PhD thesis publicly documented an earlier analytical context related to the Drosophila SARM1/ARM binding site, integrating the 60-state alignment framework with NMR-related observations.