BayesDeBulk: A Flexible Bayesian Algorithm for the Deconvolution of Bulk Tumor Data

Abstract

Characterizing the tumor microenvironment is crucial in order to improve responsiveness to immunotherapy and develop new therapeutic strategies. The fraction of different cell-types in the tumor microenvironment can be estimated based on transcriptomic profiling of bulk tumor data via deconvolution algorithms. One class of such algorithms, known as reference-based, rely on a reference signature containing gene expression data for various cell-types. The limitation of these methods is that such a signature is derived from the gene expression of pure cell-types, which might not be consistent with the transcriptomic profiling in solid tumors. On the other hand, reference-free methods usually require only a set of cell-specific markers to perform deconvolution; however, once the different components have been estimated from the data, their labeling can be problematic. To overcome these limitations, we propose BayesDeBulk - a new reference-free Bayesian method for bulk deconvolution based on gene expression data. Given a list of markers expressed in each cell-type (cell-specific markers), a repulsive prior is placed on the mean of gene expression in different cell-types to ensure that cell-specific markers are upregulated in a particular component. Contrary to existing reference-free methods, the labeling of different components is decided a priori through a repulsive prior. Furthermore, the advantage over reference-based algorithms is that the cell fractions as well as the gene expression of different cells are estimated from the data, simultaneously. Given its flexibility, BayesDeBulk can be utilized to perform bulk deconvolution beyond transcriptomic data, based on other data types such as proteomic profiles or the integration of both transcriptomic and proteomic profiles.

Pan-cancer proteogenomics characterization of tumor immunity

“…To gain insights into the infiltration pattern of different immune/stromal cell types in these tumors, we estimated cell type composition fractions in the tumor microenvironment (TME) using a recently developed deconvolution algorithm, 7 which leverages matched bulk gene expression and proteomic profiles to perform tissue deconvolution ( Table S1 ). The comparison of cell type fractions among different tumors revealed extensive cell type composition heterogeneity across different cancers ( Figure 1B ).…”

“…To understand tissue function, we must understand its varied composition at the cellular level. We inferred the cell type compositions of all tumor samples based on both transcriptomics and proteomicss via BayesDeBulk, 7 a deconvolution method that integrates proteogenomic data. The overall load of immune cells was linked to patient PFS outcomes in various cancers, including CCRCC, LUAD, PDAC, and CO ( Figure 1D ).…”

“…Employing novel algorithms for dissecting the tumor microenvironment accounting for both RNA and protein expression, 7 we illuminated the tumor molecular microenvironment, explored the pathways related to its function, and associated it with clinical outcomes. Despite many differences across the 10 cancers, our analysis revealed seven pan-cancer immune subtypes.…”

Gut-associated lymphoid tissue attrition associates with response to anti-α4β7 therapy in ulcerative colitis

“…compared pre-and post-VDZ in 31 patients with UC (cohort 5) (25). Using a deconvolution algorithm (26), we estimated the fraction of FB cells in responders and NRs. Consistent with our IF microscopy data, the fraction of naïve B cells was decreased in responders post-VDZ [P = 0.045, ∆ R = −0.14 (−63%)], with no significant change in NRs [P = 0.56, ∆ NR = −0.02 (8%)] (Fig.…”

Decomprolute: A benchmarking platform designed for multiomics-based tumor deconvolution

“…Pseudo-bulk data was simulated in a similar fashion as in Petralia et al (2022) 38 . Our simulation framework relied on two published datasets.…”