Citation

Ng, Grace Yee Lin and Ong, Chia Sui and Tan, Shing Chiang (2024) A Multiclass Method for Selecting Differentially-Expressed and Cell-Type-Discriminative Genes from scRNA-Seq Data. In: ICCBB '23: Proceedings of the 2023 7th International Conference on Computational Biology and Bioinformatics, December 2023.

Text 28.pdf - Published Version Restricted to Repository staff only Download (1MB)

Abstract

Log fold change (LFC) is a common measure used in differential expression analysis to examine the differences in gene expression between two experimental classes, as in the data generated by microarray or bulk RNA sequencing. Many single-cell RNA-seq (scRNA-seq) data are labelled with three or more classes in terms of cell types, cell states, or cell stages. Several differential expression methods have been introduced to select differentially expressed genes (DEGs) among different classes in scRNA-seq data while accounting for the technical and biological variations. However, these methods are only applicable to perform comparisons between two classes. Methods to select DEGs with multiclass comparisons have also been introduced in the literature, but different measures are used instead of LFC. Thus, this study aims to impose the impactful LFC measure as a multiclass DEGs selection method. The majority voting concept is incorporated to aggregate the DEGs from every pairwise class comparison. Cell type classification using the selected genes has been performed to evaluate and validate the genes selected by the multiclass LFC method. The results show that the proposed method is capable of reducing the number of genes to as low as 26.05% of the initial scRNA-seq data. Moreover, the selected genes can classify cells into their respective cell types more accurately (an accuracy of 0.9425) as compared to the existing scRNA-seq gene selection method (an accuracy of 0.9137).

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	multiclass, log fold change, single-cell RNA sequencing, gene selection, cell-type identification
Subjects:	Q Science > QH Natural history > QH426 Genetics
Divisions:	Faculty of Information Science and Technology (FIST)
Depositing User:	Ms Nurul Iqtiani Ahmad
Date Deposited:	27 Mar 2024 00:17
Last Modified:	27 Mar 2024 00:17
URII:	http://shdl.mmu.edu.my/id/eprint/12187

Downloads

Edit (login required)