Toward Convex Manifolds: A Geometric Perspective for Deep Graph Clustering of Single-cell RNA-seq Data

The deep clustering paradigm has shown great potential for discovering complex patterns that can reveal cell heterogeneity in single-cell RNA sequencing data. This paradigm involves two training phases: pretraining based on a pretext task and fine-tuning using pseudo-labels. Although current models yield promising results, they overlook the geometric distortions that regularly occur during the training process. More precisely, the transition between the two phases results in a coarse flattening of the latent structures, which can deteriorate the clustering performance. In this context, existing methods perform euclidean-based embedding clustering without ensuring the flatness and convexity of the latent manifolds. To address this problem, we incorporate two mechanisms. First, we introduce an overclustering loss to flatten the local curves. Second, we propose an adversarial mechanism to adjust the global geometric configuration. More precisely, the second mechanism gradually transforms the latent structures into convex ones. Empirical results on a variety of gene expression datasets show that our model outperforms state-of-the-art methods.

Architecture

The neural network architecture of our approach as defined in scTCM.py

Requirements

Installing the requirements using pip

$ pip install -r requirements.txt

#Download data The link to the datasets: https://drive.google.com/drive/folders/1fgsoyOFo5G2tKZXxLfbMrV850RM_VuqF?usp=share_link

Usage

To evaluate our approach on all datasets, run the Python scripts main_$dataset_name$.py.

$ python main_Young.py  

Arguments

The values of three hyperparameters depend on the dataset:

• threshold_2: The clustering threshold β_c
• threshold_3: The overclustering threshold β_o
• overclustering: The number of overclusters n_o

Citation

@inproceedings{ijcai2023p540, title = {Toward Convex Manifolds: A Geometric Perspective for Deep Graph Clustering of Single-cell RNA-seq Data}, author = {Mrabah, Nairouz and Amar, Mohamed Mahmoud and Bouguessa, Mohamed and Diallo, Abdoulaye Banire}, booktitle = {Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence, {IJCAI-23}}, publisher = {International Joint Conferences on Artificial Intelligence Organization}, editor = {Edith Elkind}, pages = {4855--4863}, year = {2023}, month = {8}, note = {Main Track}, doi = {10.24963/ijcai.2023/540}, url = {https://doi.org/10.24963/ijcai.2023/540}, }