Likelihood-based deconvolution of bulk gene expression data using single-cell references

Erdmann-Pham, Dan D.; Fischer, Jonathan; Hong, Justin; Song, Yun S.

doi:10.1101/gr.272344.120

Cited by 27 publications

(29 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For deconvolution, we utilized RNA-Sieve (v0.1.4) 40 . Once an integrated scRNA-Seq dataset was properly annotated by cell type, we subset this to include only the 4 olfactory mucosal biopsies from Durante et al, 2020 in order to streamline computational workflow.…”

Section: Methodsmentioning

confidence: 99%

“…We eliminated genes only present in either the scRNA-Seq dataset or the bulk RNA-Seq datasets using index.intersection(). Raw counts for scRNA-Seq and bulk RNA-Seq datasets were prepped for running the deconvolution model with default settings, as described 40 . The deconvolution model was trained by running model_from_raw_counts() on the processed scRNA-Seq reference dataset.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Deconstructing olfactory epithelium developmental pathways in esthesioneuroblastoma

Finlay

Hachem

Jang

et al. 2022

Preprint

View full text Add to dashboard Cite

ImportanceEsthesioneuroblastoma (ENB) is a rare tumor arising from the olfactory cleft region of the nasal cavity. Due to the low incidence of this tumor, as well as an absence of established reagents such as cell lines or murine models, understanding the mechanisms driving ENB pathobiology has been challenging.ObjectiveHere, we sought to apply advances from research on the human olfactory epithelial neurogenic niche, along with new biocomputational approaches, to better understand the cellular and molecular factors in low and high grade ENB and how specific transcriptomic markers may predict prognosis.Design, Setting, and ParticipantsThis was a retrospective biocomputational analysis utilizing human ENB and normal olfactory mucosal tissue and clinical outcomes data. The setting was a large academic medical center. Participants were selected based on available datasets, and included ENB patients across all four tumor grades, and patients with normal olfactory mucosa.Main Outcomes and MeasuresOutcomes include deconvolution analysis, differential gene expression analysis, overall survival, and immunohistochemistry. A machine learning model was used to predict cell-type proportion signatures in ENB, and Kaplan-Meier curves with associated log-rank tests were used to estimate differences in overall survival.ResultsWe analyzed a total of 19 ENB samples (9 low grade I/II, 10 high grade III/IV) with available bulk RNA-Sequencing and survival data, along with 10 samples from normal olfactory mucosa (3 bulk RNA-Sequencing, 7 single cell RNA-Sequencing). The bulk RNA-Sequencing deconvolution model identified a significant increase in globose basal cell (GBC) and CD8 T cell identities in high grade tumors (GBC from approximately 0% to 8%, CD8 T cell from 0.7% to 2.2%), and significant decreases in mature neuronal, Bowman’s gland, and olfactory ensheathing programs, in high grade tumors (mature neuronal from 3.7% to approximately 0%, Bowman’s gland from 18.6% to 10.5%, olfactory ensheathing from 3.4% to 1.1%). Trajectory analysis identified potential regulatory pathways in proliferative ENB cells, including PRC2. Survival analysis guided by gene expression in bulk RNA-Sequencing data identified favorable prognostic markers such as SOX9, S100B, and PLP1 expression.Conclusions and RelevanceOur analyses provide a basis for additional translational research on ENB management, as well as identification of potential new prognostic markers.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Deconstructing olfactory epithelium developmental pathways in esthesioneuroblastoma

Finlay

Hachem

Jang

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…For deconvolution performance on the pseudo-bulk and real bulk data with ground truth, we benchmark Scaden, RNAsieve, CIBER-SORTx, DWLS, MuSiC and Bisque [8][9][10][11][12]14 . We will describe the details of the benchmarking procedure below.…”

Section: Software Comparison and Settingsmentioning

confidence: 99%

Deep autoencoder for interpretable tissue-adaptive deconvolution and cell-type-specific gene analysis

et al. 2022

View full text Add to dashboard Cite

Single-cell RNA-sequencing has become a powerful tool to study biologically significant characteristics at explicitly high resolution. However, its application on emerging data is currently limited by its intrinsic techniques. Here, we introduce Tissue-AdaPtive autoEncoder (TAPE), a deep learning method connecting bulk RNA-seq and single-cell RNA-seq to achieve precise deconvolution in a short time. By constructing an interpretable decoder and training under a unique scheme, TAPE can predict cell-type fractions and cell-type-specific gene expression tissue-adaptively. Compared with popular methods on several datasets, TAPE has a better overall performance and comparable accuracy at cell type level. Additionally, it is more robust among different cell types, faster, and sensitive to provide biologically meaningful predictions. Moreover, through the analysis of clinical data, TAPE shows its ability to predict cell-type-specific gene expression profiles with biological significance. We believe that TAPE will enable and accelerate the precise analysis of high-throughput clinical data in a wide range.

show abstract

“…For deconvolution performance on the pseudo-bulk and real bulk data with ground truth, we benchmark Scaden, RNAsieve, CIBERSORTx, DWLS, MuSiC and Bisque [8][9][10][11][12]14]. We will describe the details of the benchmarking procedure below.…”

Section: Software Comparison and Settingsmentioning

confidence: 99%

“…The existing methods can be roughly divided into two categories: statistical learning-based and deep learning-based methods. Based on traditional regression models like non-negative least squares (NNLS) and support vector regression (SVR), a series of methods like CIBERSORT (CS) [7], MuSiC [8], CIBERSORTx (CSx) [9], Bisque [10], DWLS [11], RNA-Sieve [12], and BLADE [13] have been developed. All these tools need a pre-selected cell-type-specific gene expression profile (GEP) or allocating different weights to different genes based on statistic value ( e .…”

Section: Introductionmentioning

confidence: 99%

Deep autoencoder for interpretable tissue-adaptive deconvolution and cell-type-specific gene analysis

Chen

Wang

Chen

et al. 2021

Preprint

View full text Add to dashboard Cite

Single-cell RNA-sequencing (RNA-seq) has become a powerful tool to study biologically significant characteristics at explicitly high resolution. However, its application on emerging data is currently limited by its intrinsic techniques. Though many methods have been proposed to analyze bulk data using single-cell profile as a reference, they are limited on the interpretability, processing speed, and data size requirement. Here, we introduce Tissue-AdaPtive autoEncoder (TAPE), a deep learning method connecting bulk RNA-seq and single-cell RNA-seq, to achieve precise prediction in short time. By constructing an interpretable decoder and training under a unique scheme, TAPE can predict cell-type fractions and cell-type-specific gene expression tissue-adaptively. Compared with existing methods on several benchmarking datasets, TAPE is more accurate (up to 40% improvement on the real bulk data) and faster. Thus, it is sensitive enough to provide biologically meaningful predictions. For example, only TAPE can predict the tendency of increasing monocytes-to-lymphocytes ratio in COVID-19 patients from mild to serious symptoms, of which estimated indices are consistent with laboratory data. More importantly, through the analysis of clinical data, TAPE shows its ability to predict cell-type-specific gene expression profiles with biological significance. Combining with single-sample gene set enrichment analysis (ssGSEA), TAPE also provides valuable clues to investigate the immune response in different virus-infected patients. We believe that TAPE will enable and accelerate the precise analysis of high-throughput clinical data in a wide range.

show abstract

Likelihood-based deconvolution of bulk gene expression data using single-cell references

Cited by 27 publications

References 35 publications

Deconstructing olfactory epithelium developmental pathways in esthesioneuroblastoma

Deconstructing olfactory epithelium developmental pathways in esthesioneuroblastoma

Deep autoencoder for interpretable tissue-adaptive deconvolution and cell-type-specific gene analysis

Deep autoencoder for interpretable tissue-adaptive deconvolution and cell-type-specific gene analysis

Contact Info

Product

Resources

About