This repo contains all scripts for testing models in PhyloClustering.jl. All scripts in Python and Julia are writing in Jupyter notebook.
We use random-sample-tree.ipynb to randomly choose the tree structures for 8-taxon and 16-taxon trees. The scripts 4-taxa-tree-simulate.ipynb, 8-16-trees-simulate.ipynb, and nni-move-check.ipynb contain all codes used to generate simulated phylogenetic trees and embed them into Eucliean space. We set seeds for simulations so it can generate consistent trees. The paths in scripts are working on certain file structures so you may need to set the path by youself.
The scripts all-taxa-check.ipynb, hierarchical-clustering.ipynb, and nni-move-check.ipynb contain all codes used to produce the accuracy of clustering results. You also need to set the paths by your self.
The scripts hc-nni-heatmap.ipynb and plot-heatmap.ipynb contain codes to plot the accuracy heatmap. You also need to set the paths by your self.
baobabs-test.ipynb contains the code to cluster trees in Baobabs dataset. We also use reroot-tree.R to reroot the trees before use Densitree to visualize them.
NNI-tree-templates.trees: 8-taxon and 16-taxon tree templates we used to simulate trees in NNI tests.
16-taxa.trees: 8-taxon and 16-taxon tree templates we used to simulate trees in all other tests.
4-taxa-tree-simulate.ipynb: Simulate 4-taxon trees and 4-taxon network in bipartition format. Writes simulated trees into csv files.
8-16-trees-simulate.ipynb: Simulate 8-taxon and 16-taxon trees in bipartition format. Writes simulated trees into csv or jld files.
all-taxa-check.ipynb: Test repeating K-means and GMM on trees with all taxon number. Writes the clustering results into csv files.
baobabs-test.ipynb: Test K-means and hierarchical clustering on Baobabs dataset.
bipartition-embedding-julia.ipynb: Transform trees from Newick format to bipartition format. Writes trees into csv files.
bipartition-to-tree.ipynb: Calculate the mean tree of the cluster. Transforms 4-taxon trees in bipartition fromat to unrooted trees.
clustering-julia.ipynb: Check if models work in Julia. Implements k-means, GMM, and DBSCAN.
dog-data-test.ipynb(Deprecated): Test K-means on Canis dataset.
hierarchical-clustering.ipynb: Implement and tests hierarchical clustering.
imbalance-and-comparison.ipynb：Test split-weight embedding on imbalanced dataset, compare it with RF distance in terms of running time, memory cost, and accuracy.
nni-move-check.ipynb: Test K-means and hierarchical clustering on trees with 0-5 NNIs.
number-to-name.ipynb(Deprecated): Change names of taxa from numbers to Canis species.
random-sample-tree.ipynb: Randomly choose 15 8-taxon trees and 15 16 taxon trees.
tree-investigation.ipynb: Investigate the visualization of the 16-taxon trees to determine if they are indeed separate clusters.
bipartition-embedding.ipynb: Transform trees from Newick format into bipartition format using the same algorithm in bipartition-embedding-julia.ipynb. Also check the performance of the algorithm.
clustering-visualize.ipynb*: Test models (K-mean, GMM, and spectral clustering) and dimensionality reduction algorithms (PCA and t-SNE) for visulazation.
clutering-check-python.ipynb: Check performance of models (K-mean, GMM, and spectral clustering)
data-clean.ipynb(Deprecated): Change names of taxa from Canis species to numbers.
final-figure.ipynb: Plot all the figures in the paper.
hc-nni-heatmap.ipynb: Visualize the results of hierarchical clustering and NNI tests.
plot-heatmap.ipynb: Visualize the results of K-means and GMM.
print-bipartition.ipynb: Transform trees from Newick format to bipartition format. Writes trees into csv files.
reroot-tree.R: Reroot trees from Canis and Baobabs datasets.