Inference of immune cell composition on the expression profiles of mouse tissue

Abstract

Mice are some of the widely used experimental animal models for studying human diseases. Defining the compositions of immune cell populations in various tissues from experimental mouse models is critical to understanding the involvement of immune responses in various physiological and patho-physiological conditions. However, non-lymphoid tissues are normally composed of vast and diverse cellular components, which make it difficult to quantify the relative proportions of immune cell types. Here we report the development of a computational algorithm, ImmuCC, to infer the relative compositions of 25 immune cell types in mouse tissues using microarray-based mRNA expression data. The ImmuCC algorithm showed good performance and robustness in many simulated datasets. Remarkable concordances were observed when ImmuCC was used on three public datasets, one including enriched immune cells, one with normal single positive T cells, and one with leukemia cell samples. To validate the performance of ImmuCC objectively, thorough cross-comparison of ImmuCC predicted compositions and flow cytometry results was done with in-house generated datasets collected from four distinct mouse lymphoid tissues and three different types of tumor tissues. The good correlation and biologically meaningful results demonstrate the broad utility of ImmuCC for assessing immune cell composition in diverse mouse tissues under various conditions.

Data-driven detection of subtype-specific and differentially expressed genes

“…where subscript ( ) indicates the subtype with the maximum mean among all subtypes. Note that OVE-FC has previously been proposed for multiclass classification 13,14 , and matches well the definition of SDEGs 5,8,19,25 . Conceptually, the null hypothesis for non-SDEGs, and the alternative hypothesis for SDEGs, can be described as non-SDEG : = 0;…”

“…In this case, the null hypothesis is that samples in all subtypes are drawn from the same population, resulting in the selection of genes that may not conform to the SDEG definition. One-Versus-Rest Fold Change (OVR-FC) is another popular method based on the ratio of the average expression in a particular subtype to that of the average expression in the rest samples [10][11][12] , and OVR t-test is occasionally used to assess the statistical significance of detected genes 13 . However, a gene with low average expression in the rest is not necessarily expressed at a low level in every subtype in the rest.…”

Latent Mycobacterium tuberculosis infection provides protection for the host by changing the activation state of the innate immune system

“…To assess more broadly whether LTBI changes the relative proportions of circulating immune cells, we performed whole blood RNA sequencing on LTBI mice and controls and applied a previously published transcriptome deconvolution approach based on a published deconvolution matrix 21,22 .…”

Late B lymphocyte action in dysfunctional tissue repair following kidney injury and transplantation

“…3f). 29,30 Beyond the sixth month after injury, lymphocytes were mostly organized in large cellular clusters, primarily around small arteries, between renal tubules identified by Havcr1 and Krt20 activity, markers of unresolved tubular injury ( Fig. 3g, Suppl.…”