We propose a novel approach that explores the interaction between local and global latent configurations to progressively adjust the reconstruction and embedding clustering tasks. We elaborate a topological and probabilistic filter to mitigate Feature Randomness and a cell-cell graph structure and content correction mechanism to counteract Feature Drift. The Zero-Inflated Negative Binomial model is also integrated to capture the characteristics of gene expression profiles. We conduct detailed experiments on eight real-world datasets from multiple representative genome sequencing platforms. Our approach outperforms the state-of-the- art clustering methods in various evaluation metrics.

The neural network architecture of our approach, the pretraining, the clustering training as well as the topological and probabilistic filter are defined in sctpf.py.

Installing the requirements using pip

$ pip install -r requirements.txt

Compile the C++ code for computing the largest connected components. If your python version is not 3.8, please modify the command inside compile_pers_lib.sh.

$ cd ref
$ ./compile_pers_lib.sh

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

To evaluate our approach on each of the eight datasets, run the python scripts main_$dataset_name$.py. The following is a list of all the scripts:

main_Muraro.py; main_Plasschaert.py; main_Qx_LM.py; main_QS_Diaphragm.py; main_QS_Heart.py; main_QS_LM.py ;main_Wang_Lung.py; main_Young.py

$ python main_Muraro.py  --k_cc=2 --threshold_2=0.7

The best hyperparameters are used by default.

The topological and probabilistic filter rely on two hyperparameters:

• threshold_2: The threshold over which the soft assignment is considered with high confidence(epsilon in Algorithm 1 of the paper)

• k_cc: The parameter used to build the k-nn graph(k in Algorithm 1 of the paper )

@inproceedings{mrabah2023exploring, title={Exploring the interaction between local and global latent configurations for clustering single-cell RNA-seq: a unified perspective}, author={Mrabah, Nairouz and Amar, Mohamed Mahmoud and Bouguessa, Mohamed and Diallo, Abdoulaye Banire}, booktitle={Proceedings of the AAAI Conference on Artificial Intelligence}, volume={37}, number={8}, pages={9235--9242}, year={2023} }