A principal feature analysis

Breitenbach, Tim; Rasbach, Lauritz; Liang, Chunguang; Jahnke, Patrick

doi:10.1016/j.jocs.2021.101502

Cited by 5 publications

(32 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Next, the preprocessed data is analyzed using the PFA 4 to identify the genes required to distinguish between the cell types (labels). Since the preprocessed data has a signi cant number of genes, this step is also rather time-consuming (approximately one and a half day in mean for our experiments and used hardware mentioned above).…”

Section: Methodsmentioning

confidence: 99%

“…The genes returned by the PFA are ranked by the mutual information they share with the label, as explained in 4 . The more mutual information an expression of a gene and a label have in common, the more information about the label can be derived from the measurement of the corresponding gene.…”

Section: Methodsmentioning

confidence: 99%

“…To solve the task of removing genes that are functions of others and select only those genes from the remaining genes that are necessary for identifying the respective cell type, we use the principal feature analysis (PFA) 4 . This method generates a graph where each node represents a gene.…”

Section: Introductionmentioning

confidence: 99%

“…For the task of gene or feature selection, respectively, there are other methods available (details in 11,4 ). However, these methods have some limitations.…”

Section: Introductionmentioning

confidence: 99%

“…Furthermore, a method purely based on mutual information also provides genes that might be redundant. On the other hand, the minimize redundancy / maximize relevance method suffers from a longer calculation time than the PFA, as elaborated in 4 . The sparse learning methods (described in detail in 11 ) use a speci c model to select genes.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Focused single-cell analysis with principal feature analysis, mutual information, and machine learning reveals cell type signatures

Breitenbach

Caliskan

Çalışkan

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

Our single-cell analysis pipeline identifies a set of key genes whose expression describes the difference between cell types and consists of (i) R packages Seurat and DoubletFinder preprocess single cell data according to quality. Next, a custom-made python script randomizes obtained data, generating random sets for comparison. (ii) Principal feature analysis (PFA) identifies genes contributing to characteristic differences in cell types, represented by corresponding labels, removing redundant genes that are functions of others, leaving only genes that are arguments of such functions. (iii) Genes associated with predicting the labels of the cells are selected. For further reduction, selected genes are ranked by the mutual information that the expression of these genes shares with the labels representing the cell types. (iv) Importantly, selected genes are validated by a machine learning framework where the prediction accuracy on all genes vs. the selected genes is tested. (v) If desired, mutual information ranking balances the number of selected genes and information content for predicting the cell types. Results from different single-cell data include: comparisons of adipocytes and adipose stem and progenitor cells (ASPCs), adipocytes and mesothelium, and endothelial cells and mesothelium, as well as fibroblast and macrophage. In each, only about ten out of more than 30000 genes are identified carrying the relevant information classifying the single cells for distinguishable cell types. Our code is provided in a GitHub repository at … ( link will be added after acceptance).

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

“…For the task of gene or feature selection, respectively, there are other methods available (details in 11,4 ). However, these methods have some limitations.…”

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations