Many computational methods have been developed to infer cell type proportions from bulk transcriptomics data. However, an evaluation of the impact of data transformation, pre-processing, marker selection, cell type composition and choice of methodology on the deconvolution results is still lacking. Using five single-cell RNA-sequencing (scRNA-seq) datasets, we generate pseudo-bulk mixtures to evaluate the combined impact of these factors. Both bulk deconvolution methodologies and those that use scRNA-seq data as reference perform best when applied to data in linear scale and the choice of normalization has a dramatic impact on some, but not all methods. Overall, methods that use scRNA-seq data have comparable performance to the best performing bulk methods whereas semi-supervised approaches show higher error values. Moreover, failure to include cell types in the reference that are present in a mixture leads to substantially worse results, regardless of the previous choices. Altogether, we evaluate the combined impact of factors affecting the deconvolution task across different datasets and propose general guidelines to maximize its performance.

Summary: Gene expression analyses of bulk tissues often ignore cell type composition as an important confounding factor, resulting in a loss of signal from lowly abundant cell types. In this review, we highlight the importance and value of computational deconvolution methods to infer the abundance of different cell types and/or cell type-specific expression profiles in heterogeneous samples without performing physical cell sorting. We also explain the various deconvolution scenarios, the mathematical approaches used to solve them and the effect of data processing and different confounding factors on the accuracy of the deconvolution results. Supplementary information: Supplementary data are available at Bioinformatics online.

Rationale: Aberrant expression of microRNAs (miRNAs) can have a detrimental role in disease pathogenesis. Objectives: To identify dysregulated miRNAs in lung tissue of patients with chronic obstructive pulmonary disease (COPD). Methods: We performed miRNA and mRNA profiling using high throughput stem-loop reverse-transcriptase quantitative polymerase chain reaction and mRNA microarray, respectively, on lung tissue of 30 patients (screening cohort) encompassing 8 never-smokers, 10 smokers without airflow limitation, and 12 smokers with COPD. Differential expression of miRNA-218-5p (miR-218-5p) was validated by reverse-transcriptase quantitative polymerase chain reaction in an independent cohort of 71 patients, an in vivo murine model of COPD, and primary human bronchial epithelial cells. Localization of miR-218-5p was assessed by in situ hybridization. In vitro and in vivo perturbation of miR-218-5p combined with RNA sequencing and gene set enrichment analysis was used to elucidate its functional role in COPD pathogenesis. Measurements and Main Results: Several miRNAs were differentially expressed among the different patient groups. Interestingly, miR-218-5p was significantly down-regulated in smokers without airflow limitation and in patients with COPD compared with never-smokers. Decreased pulmonary expression of miR-218-5p was validated in an independent validation cohort, in cigarette smoke–exposed mice, and in human bronchial epithelial cells. Importantly, expression of miR-218-5p strongly correlated with airway obstruction. Furthermore, cellular localization of miR-218-5p in human and murine lung revealed highest expression of miR-218-5p in the bronchial airway epithelium. Perturbation experiments with a miR-218-5p mimic or inhibitor demonstrated a protective role of miR-218-5p in cigarette smoke–induced inflammation and COPD. Conclusions: We highlight a role for miR-218-5p in the pathogenesis of COPD.