Comparison of Methods for Feature Selection in Clustering of High-Dimensional RNA-Sequencing Data to Identify Cancer Subtypes

Abstract

Cancer subtype identification is important to facilitate cancer diagnosis and select effective treatments. Clustering of cancer patients based on high-dimensional RNA-sequencing data can be used to detect novel subtypes, but only a subset of the features (e.g., genes) contains information related to the cancer subtype. Therefore, it is reasonable to assume that the clustering should be based on a set of carefully selected features rather than all features. Several feature selection methods have been proposed, but how and when to use these methods are still poorly understood. Thirteen feature selection methods were evaluated on four human cancer data sets, all with known subtypes (gold standards), which were only used for evaluation. The methods were characterized by considering mean expression and standard deviation (SD) of the selected genes, the overlap with other methods and their clustering performance, obtained comparing the clustering result with the gold standard using the adjusted Rand index (ARI). The results were compared to a supervised approach as a positive control and two negative controls in which either a random selection of genes or all genes were included. For all data sets, the best feature selection approach outperformed the negative control and for two data sets the gain was substantial with ARI increasing from (−0.01, 0.39) to (0.66, 0.72), respectively. No feature selection method completely outperformed the others but using the dip-rest statistic to select 1000 genes was overall a good choice. The commonly used approach, where genes with the highest SDs are selected, did not perform well in our study.

Entropy sorting of single-cell RNA sequencing data reveals the inner cell mass in the human pre-implantation embryo

“…HVG selection seeks to identify a subset of genes more predictive of distinct cell types than randomly expressed genes. While it is a widely used pre-processing strategy, HVG selection can struggle to account for important but lowly expressed genes or genes present in only a small fraction of cells ( Källberg et al., 2021 ). Furthermore, evaluation of various HVG methods found that different techniques show poor overlap in HVGs suggested from the same datasets and that highly expressed genes were often incorrectly flagged as HVGs ( Yip et al., 2018 ).…”

Entropy sorting of single cell RNA sequencing data reveals the inner cell mass in the human pre-implantation embryo

“…HVG selection seeks to identify a subset of genes more predictive of distinct cell types than randomly expressed genes. While it is a widely used pre-processing strategy, HVG selection can struggle to account for important but lowly expressed genes, or genes present in only a small fraction of cells (Källberg, Vidman, and Rydén 2021). Furthermore, evaluation of various HVG methods found that different techniques show poor overlap in HVG suggested from the same datasets, and that highly expressed genes were often incorrectly flagged as HVGs (Yip, Sham, and Wang 2018).…”

Novel feature selection methods for construction of accurate epigenetic clocks

“…Given the vast feature space of -omics datasets, creating an accurate model through penalised regression is often not difficult, however finding the right features for further study to infer biological understanding is harder. In recent years, feature selection has become a popular method for novel biomarker discovery [ 35 , 36 , 37 , 38 , 39 , 40 ] and the application of the novel feature selection methods in this paper could accelerate the discovery of biomarkers in many fields.…”