Fuzzy C-Means based LSED analysis model of single-cell RNA sequencing data for gene expression difference data

Abstract

Analysis of single-cell RNA sequencing (scRNA-seq) data is often complicate due to the sparsity and high data dimensionality. In this work, we proposed Fuzzy C-means based linear stable-exponential distribution (LSED) model for analyzing scRNA-seq count data of chronic myeloid leukemia (CML) patients. We propose pipelines stages for analysis in which noisy and inconsistent data form sequencing is removed during data preprocessing, this process data then form the cluster of gene feature using fuzzy c-means (FCM) clustering, relevant features are extracted during feature extraction approach. These extracted features are then fed into LSED model in order to difference feature data of gene expression. Finally we evaluate the performance for proposed analysis model based on parameter estimation, distribution comparison and parameter analysis. From the result analysis it was observed that proposed analysis model parameter reflect change in condition of patient more effectively as well as this model fits difference data of gene expression in more better way in comparison to Cauchy and stable distribution. Additional, the results of Gene-set enrichment analysis specify the affinity of proposed model can replicate the distinct enhancement of BCR-ABL+ stem cell as well as BCR-ABL- stem cells. Significantly, proposed FCM based LSED analysis model studies CML from the perspective of statistical models, which present a new sight for CML scientific research.