Journal: bioRxiv

Article Title: Deciphering context-dependent epigenetic program by network-based prediction of clustered open regulatory elements from single-cell chromatin accessibility

doi: 10.64898/2026.03.17.712366

Figure Lengend Snippet: a Chow-Ruskey diagram of Non-CORE Active Enhancer (non-CORE)-associated genes and Clustered Open Regulatory Element (CORE)-associated genes in PBMC. B cells (red border), Monocytes (purple border), T cells (green border), NK cells (brown border). Monocytes represent the union of CORE-associated genes derived from CD14+ Monocytes and CD16+ Monocytes, while T cells represent the combined set of CORE-associated genes derived from CD4 Memory T cells and CD8 Effector T cells. The color of the borders around each intersection corresponds to the cell types whose genes overlap. The area of each intersection is proportional to the number of genes within the intersection. b Proportion of CORE constituents that overlap with H3K4me1 peaks, relative to the total number of enhancer candidates in each cell type. NK: NK cells, CD14_Mono: CD14+ Monocytes, B: B cells, CD8_Teff: CD8+ Effector T cells, CD4_Tmem: CD4+ Memory T cells. c Heatmap for cell type-specific master regulators. Black: gene present in the CORE-associated genes of the given cell type. White: absent. d Log2-transformed fold changes of normalized TPM of genes in each cell type. Normalized TPM values were retrieved from the PBMC Monaco dataset in the Human Protein Atlas (HPA). Left: comparison between typical enhancer (TE)-associated genes and super-enhancer (SE)-associated genes. Right: comparison between non-CORE-associated genes and CORE-associated genes. P-value was calculated by a two-sided Wilcoxon signed-rank test. e Representative RNA expression for GATA3 , EBF1 , and MAFB . Normalized TPM values were retrieved from the HPA PBMC dataset. CD4 T Mem: CD4+ Memory T cells, CD8 T Mem: CD8+ Memory T cells, NK: NK cells, B mem: Memory B cells, B nai: naïve B cells, cDC: classical Dendritic cells, CD14 Mono: CD14+ Monocytes, CD16 Mono: CD16+ Monocytes. f Enrichment analysis of known TF motifs within CORE and non-CORE constituents across cell types. Colored by enrichment score, modified MinMax-scaled statistical significance (-log10(P-value)).

Article Snippet: To investigate cell type-specific transcriptional regulation of CORE-associated and SE-associated genes in PBMC, we utilized the Monaco PBMC RNA-seq dataset and the HPA PBMC RNA-seq dataset from the Human Protein Atlas (HPA; https://www.proteinatlas.org/ ) [ ].

Techniques: Derivative Assay, Transformation Assay, Comparison, RNA Expression, Modification

Journal: bioRxiv

Article Title: Deciphering context-dependent epigenetic program by network-based prediction of clustered open regulatory elements from single-cell chromatin accessibility

doi: 10.64898/2026.03.17.712366

Figure Lengend Snippet: a Chow-Ruskey diagram of Non-CORE Active Enhancer (non-CORE)-associated genes and Clustered Open Regulatory Element (CORE)-associated genes in PBMC. B cells (red border), Monocytes (purple border), T cells (green border), NK cells (brown border). Monocytes represent the union of CORE-associated genes derived from CD14+ Monocytes and CD16+ Monocytes, while T cells represent the combined set of CORE-associated genes derived from CD4 Memory T cells and CD8 Effector T cells. The color of the borders around each intersection corresponds to the cell types whose genes overlap. The area of each intersection is proportional to the number of genes within the intersection. b Chow-Ruskey diagram of typical enhancer (TE)-associated genes and super-enhancer (SE)-associated genes. B cells (red border), Monocytes (purple border), T cells (green border), NK cells (brown border). Monocytes represent the union of SE-associated genes derived from CD14+ Monocytes and CD16+ Monocytes, while T cells represent the combined set of SE-associated genes derived from CD4+ T cells and CD8+ T cells. The color of the borders around each intersection corresponds to the cell types whose genes overlap. The area of each intersection is proportional to the number of genes within the intersection. c Log2-transformed fold changes of normalized TPM of genes in each cell type. Normalized TPM values were retrieved from the PBMC Monaco dataset in the Human Protein Atlas (HPA). Comparison between non-CORE-associated genes and CORE-associated genes using the potential option. P-value was calculated by a two-sided Wilcoxon signed-rank test. d Dot plot showing the classification performance of CORE in distinguishing SE and TE. Dots are colored by Area Under Curve (AUC), and the dot sizes indicate the specificity. SE derived from B: B cells, CD14: CD14+ Monocytes, CD16: CD16+ Monocytes, CD4: CD4+ T cells, CD4M: CD4+ Memory T cells, CD8: CD8+ T cells, NK: NK cells. e Representative ROC curves from CD16+ Monocytes. Left: CORE from the potential option, Right: CORE from the active option. The AUC values are indicated in the lower right corner of each plot.

Article Snippet: To investigate cell type-specific transcriptional regulation of CORE-associated and SE-associated genes in PBMC, we utilized the Monaco PBMC RNA-seq dataset and the HPA PBMC RNA-seq dataset from the Human Protein Atlas (HPA; https://www.proteinatlas.org/ ) [ ].

Techniques: Derivative Assay, Transformation Assay, Comparison

Journal: bioRxiv

Article Title: Deciphering context-dependent epigenetic program by network-based prediction of clustered open regulatory elements from single-cell chromatin accessibility

doi: 10.64898/2026.03.17.712366

Figure Lengend Snippet: a UMAP embedding of the PBMC scCUT&Tag-seq dataset, colored and annotated by clusters. b UMAP embedding of the PBMC scCUT&Tag-seq dataset, colored and annotated by cell types. CD14_Mono: CD14+ Monocytes, CD4_Tmem: CD4+ Memory T cells, CD8_Teff: CD8+ Effector T cells, B: B cells, NK: NK cells. c Chow-Ruskey diagram of typical enhancer (TE)-associated genes and super-enhancer (SE)-associated genes. B cells (red border), Monocytes (purple border), T cells (green border), NK cells (brown border). Monocytes represent the union of SE-associated genes derived from CD14+ Monocytes, while T cells represent the combined set of SE-associated genes derived from CD4+ Memory T cells and CD8+ Effector T cells. The color of the borders around each intersection corresponds to the cell types whose genes overlap. The area of each intersection is proportional to the number of genes within the intersection. d Heatmap for cell type-specific master regulators. Black: gene present in the SE-associated genes of the given cell type. White: absent. e Log2-transformed fold changes of normalized TPM of genes in each cell type. Normalized TPM values were retrieved from the PBMC Monaco dataset in the Human Protein Atlas (HPA). Comparison between TE-associated genes and SE-associated genes. P-value was calculated by a two-sided Wilcoxon signed-rank test.

Article Snippet: To investigate cell type-specific transcriptional regulation of CORE-associated and SE-associated genes in PBMC, we utilized the Monaco PBMC RNA-seq dataset and the HPA PBMC RNA-seq dataset from the Human Protein Atlas (HPA; https://www.proteinatlas.org/ ) [ ].

Techniques: Derivative Assay, Transformation Assay, Comparison

Journal: bioRxiv

Article Title: Deciphering context-dependent epigenetic program by network-based prediction of clustered open regulatory elements from single-cell chromatin accessibility

doi: 10.64898/2026.03.17.712366

Figure Lengend Snippet: a NMF analysis of H3K27ac ChIP-seq signals within CORE to distinguish between CRC and Normal samples. Dots are colored by disease state: CRC (red color), Normal (blue color). b Log2-transformed fold changes of H3K27ac ChIP-seq signals between CRC and Normal within CORE, SE, and TE. Dots and borderlines are colored by region categories: CORE (red color), SE (blue color), TE (grey color). c Enrichment analysis of de novo TF motifs in CORE and non-CORE constituents. Left: TF motif enrichment analysis using non-CORE constituents, Right: TF motif enrichment analysis using CORE constituents. The FOXM1 gene is highlighted in red. d TCGA survival analysis for MACC1 and EGFR genes. P-values were calculated using the Log-rank test. Lines are colored by expression level status: High gene expression (red color), Low gene expression (blue color). Survival curves for the MACC1 and EGFR genes were obtained from RNA-seq and RPPA data, respectively. e Track visualization result for the USP7 gene with scATAC-seq signals normalized by Reads in TSS. Colored and annotated by disease state: CRC (Colorectal cancer epithelial cells; red color), Normal (Normal epithelial cells; blue color).

Article Snippet: To investigate cell type-specific transcriptional regulation of CORE-associated and SE-associated genes in PBMC, we utilized the Monaco PBMC RNA-seq dataset and the HPA PBMC RNA-seq dataset from the Human Protein Atlas (HPA; https://www.proteinatlas.org/ ) [ ].

Techniques: ChIP-sequencing, Transformation Assay, Expressing, Gene Expression, RNA Sequencing

Journal: bioRxiv

Article Title: Deciphering context-dependent epigenetic program by network-based prediction of clustered open regulatory elements from single-cell chromatin accessibility

doi: 10.64898/2026.03.17.712366

Figure Lengend Snippet: TCGA survival analysis for CORE-associated genes. a ANO1 , b ARL4C , c FOXD1 , d PRDM15 , e HOXC10 , f ISM1 , g BRD2 , h KDM4B . P-values were calculated using the Log-rank test. Lines are colored by expression level status: High gene expression (red color), Low gene expression (blue color). Survival curves were obtained from RNA-seq data.

Article Snippet: To investigate cell type-specific transcriptional regulation of CORE-associated and SE-associated genes in PBMC, we utilized the Monaco PBMC RNA-seq dataset and the HPA PBMC RNA-seq dataset from the Human Protein Atlas (HPA; https://www.proteinatlas.org/ ) [ ].

Techniques: Expressing, Gene Expression, RNA Sequencing

Journal: bioRxiv

Article Title: Deciphering context-dependent epigenetic program by network-based prediction of clustered open regulatory elements from single-cell chromatin accessibility

doi: 10.64898/2026.03.17.712366

Figure Lengend Snippet: a Proportion of CORE- and SE-associated genes that overlap with metastasis-related genes, relative to the total number of metastasis-related genes. Target (Bulk): differentially expressed genes between metastatic tumor and primary CRC in bulk RNA-seq data overlapping with target genes in TF regulons from scRNA-seq data, TF (scRNA): differential TFs between metastatic tumor and primary CRC based on TF regulon activity in scRNA-seq data. b Box plots showing weighted degree and weighted transitivity (weighted local clustering coefficients) in CORE only, SE only, and Other peaks. CORE only: H3K4me1 peaks located exclusively within CORE, SE only: H3K4me1 peaks located exclusively within SE, Other peaks: H3K4me1 peaks that do not fall into either CORE or SE. Correlation coefficients are used as edge weights in the co-accessibility network. c Log2-transformed fold changes of mean weighted degree and mean clustering coefficient between CORE and SE. DEGR: Weighted degree (W.D), LCLC: Weighted transitivity (W.T). d Fold changes of the number of metastasis-related genes that overlap with the annotated genes from peaks with a high weighted local clustering coefficient, relative to the number of metastasis-related genes that overlap with the annotated genes from peaks with a high weighted degree. Exp: Expected fold changes; the number of annotated genes from peaks with a high weighted clustering coefficient divided by the number of annotated genes from peaks with a high weighted degree, Obs: Observed fold changes. Target (Bulk): differentially expressed genes between metastatic tumor and primary CRC in bulk RNA-seq data overlapping with target genes in TF regulons from scRNA-seq data, TF (scRNA): differential TFs between metastatic tumor and primary CRC based on TF regulon activity in scRNA-seq data.

Article Snippet: To investigate cell type-specific transcriptional regulation of CORE-associated and SE-associated genes in PBMC, we utilized the Monaco PBMC RNA-seq dataset and the HPA PBMC RNA-seq dataset from the Human Protein Atlas (HPA; https://www.proteinatlas.org/ ) [ ].

Techniques: RNA Sequencing, Activity Assay, Transformation Assay

Journal: bioRxiv

Article Title: Deciphering context-dependent epigenetic program by network-based prediction of clustered open regulatory elements from single-cell chromatin accessibility

doi: 10.64898/2026.03.17.712366

Figure Lengend Snippet: a Track visualization result for the USP7 gene with H3K27ac HiChIP interactions from HT29 cells and normalized H3K27ac ChIP-seq signals in CRC and Normal samples. Colored and annotated by disease state: CRC (Colorectal cancer epithelial cells; red color), Normal (Normal epithelial cells; blue color). b Track visualization of RNA-seq from Left Colon ontology for in silico enhancer knockout within CORE near USP7 gene. c The magnitude of the predicted effect in RNA-seq reads from Left Colon ontology for in silico enhancer knockout within CORE near USP7 gene.

Article Snippet: To investigate cell type-specific transcriptional regulation of CORE-associated and SE-associated genes in PBMC, we utilized the Monaco PBMC RNA-seq dataset and the HPA PBMC RNA-seq dataset from the Human Protein Atlas (HPA; https://www.proteinatlas.org/ ) [ ].

Techniques: HiChIP, ChIP-sequencing, RNA Sequencing, In Silico, Knock-Out

Journal: Advanced Science

Article Title: scPER: A Rigorous Computational Approach to Determine Cellular Subtypes in Tumors Aligned With Cancer Phenotypes From Total RNA Sequencing

doi: 10.1002/advs.202514502

Figure Lengend Snippet: Overview of scPER and the function of each stage. Stage 1. Reference integration and preprocessing. scRNA‐seq datasets from multiple studies or tissues are integrated to form a reference matrix R (rows = genes, columns = cells). After QC and feature selection, denoising and imputation are applied to stabilize gene signals across datasets. Output: preprocessed single‐cell expression R with cell‐type labels and confounder labels (study/platform/tissue). Stage 2. Confounder‐aware representation learning. An adversarial autoencoder learns a low‐dimensional latent space Z from R . The encoder E maps expression x to z ; the decoder D reconstructs x from z . An adversary V tries to predict designated confounders c (e.g., study/platform/tissue) from z . Training minimizes reconstruction while implicitly maximizing adversary loss with weight λ (i.e., L rec − λ L adv ), yielding Z that preserves biology but reduces nuisance variation. Output: cell embeddings Z and a trained encoder E for embedding new data. Stage 3. Bulk embedding and proportion estimation. Bulk RNA‐seq samples T are normalized like the single‐cell data and passed through the trained encoder E to obtain bulk embeddings in the same latent space. An XGBoost regressor is trained on latent features from single‐cell references to predict cell‐type compositions P for each bulk sample (columns = samples, rows = cell types). Output: estimated cell‐type proportions P across all bulk samples. Stage 4. Downstream biology and clinical association. The estimated compositions P enable downstream analyses: (i) identify critical genes modulating the cell proportions, (ii) test clinical endpoints such as survival or response to immunotherapy, and (iii) identify phenotype‐associated changes in cell types or subclusters. R = reference single‐cell matrix; T = bulk expression; Z = scPER latent embeddings; P = estimated cell‐type proportions; λ = adversarial weight; c = confounder label. Arrows indicate data flow from reference construction to embedding, deconvolution, and phenotype analyses.

Article Snippet: In this study, the PBMC scRNA‐seq dataset in Figure was downloaded from the 10X Genomics data download page: 8k PBMCs from a healthy donor ( https://www.10xgenomics.com/resources/datasets/8‐k‐pbm‐cs‐from‐a‐healthy‐donor‐2‐standard‐2‐1‐0 ).

Techniques: Selection, Single Cell, Expressing, RNA Sequencing

Journal: Advanced Science

Article Title: scPER: A Rigorous Computational Approach to Determine Cellular Subtypes in Tumors Aligned With Cancer Phenotypes From Total RNA Sequencing

doi: 10.1002/advs.202514502

Figure Lengend Snippet: Cross‑tissue deconvolution performance comparison. (A) Workflow and latent embedding of the cross‑tissue reference panel. Left: UMAP of single‑cell transcriptomes from melanoma, ovarian cancer ascites, and PBMC sources before batch correction. Top subpanel: cells colored by tissue of origin; bottom subpanel: cells colored by annotated cell type. Middle: heatmap of scPER's 100 latent representations after adversarial batch correction, with rows as embedding features and columns as six cell types. (B) Pearson correlation coefficients between true and predicted cell‑type fractions for each deconvolution tool, summarizing overall accuracy across the 16 simulated prostate cancer bulk samples. (C‐F) True versus predicted proportion scatter plots for monocytes/macrophages using (C) scPER, (D) CIBERSORTx, (E) BayesPrism, and (F) Scaden. Points represent individual samples; the x‑axis shows ground‑truth fractions (from summed single‑cell counts), and the y‑axis shows model predictions.

Article Snippet: In this study, the PBMC scRNA‐seq dataset in Figure was downloaded from the 10X Genomics data download page: 8k PBMCs from a healthy donor ( https://www.10xgenomics.com/resources/datasets/8‐k‐pbm‐cs‐from‐a‐healthy‐donor‐2‐standard‐2‐1‐0 ).

Techniques: Comparison

Journal: Advanced Science

Article Title: scPER: A Rigorous Computational Approach to Determine Cellular Subtypes in Tumors Aligned With Cancer Phenotypes From Total RNA Sequencing

doi: 10.1002/advs.202514502

Figure Lengend Snippet: scPER‐derived cell‐type proportions predict melanoma immunotherapy outcomes. (A‐B) UMAP projections of scRNA‐seq from two melanoma datasets before scPER integration. Colored by annotated cell types. Red rectangles highlight cell populations unique to one study.(C) Boxplots comparing estimated proportions of each cell type between responders (R; n=14) and non‐responders (NR; n = 13) to immunotherapy. P‐values were calculated using the Wilcoxon rank‐sum test. (D) Receiver operating characteristic (ROC) curves for XGBoost classifier predicting response status: one using scPER‐estimated cell‐type proportion features alone and another using tumor mutation burden (mutation load) alone. Area under the ROC curve (AUROC) values are indicated for each model. (E) Kaplan‐Meier survival analysis of LTB expression in the TCGA‐SKCM cohort. Patients were stratified into high‐ and low‐ LTB expression groups based on the median; the log‐rank test p‐value is shown.

Article Snippet: In this study, the PBMC scRNA‐seq dataset in Figure was downloaded from the 10X Genomics data download page: 8k PBMCs from a healthy donor ( https://www.10xgenomics.com/resources/datasets/8‐k‐pbm‐cs‐from‐a‐healthy‐donor‐2‐standard‐2‐1‐0 ).

Techniques: Derivative Assay, Mutagenesis, Expressing

Journal: Advanced Science

Article Title: scPER: A Rigorous Computational Approach to Determine Cellular Subtypes in Tumors Aligned With Cancer Phenotypes From Total RNA Sequencing

doi: 10.1002/advs.202514502

Figure Lengend Snippet: Benchmarking deconvolution accuracy on PBMC single‐cell and bulk profiles.(A) t‐Distributed Stochastic Neighbor Embedding (t‐SNE) of the 100 latent embeddings learned from scRNA‐seq data, with points colored by the six annotated immune cell types. (B) Pearson correlation coefficients ( r ) between true and predicted cell‐type fractions across all bulk samples, shown for each deconvolution tool as a violin plot summarizing the distribution of r values. (C) True versus predicted proportion scatter plots for each cell type and method. Each point represents one bulk sample (n = 164); the x‐axis shows the ground‐truth fraction (flow cytometry), and the y‐axis shows the fraction predicted by the deconvolution tool.

Article Snippet: In this study, the PBMC scRNA‐seq dataset in Figure was downloaded from the 10X Genomics data download page: 8k PBMCs from a healthy donor ( https://www.10xgenomics.com/resources/datasets/8‐k‐pbm‐cs‐from‐a‐healthy‐donor‐2‐standard‐2‐1‐0 ).

Techniques: Single Cell, Flow Cytometry

Journal: Genome Biology

Article Title: SCAPTURE: a deep learning-embedded pipeline that captures polyadenylation information from 3′ tag-based RNA-seq of single cells

doi: 10.1186/s13059-021-02437-5

Figure Lengend Snippet: Developing SCAPTURE to identify cleavage and polyadenylation sites (PASs) from 3′ tag-based scRNA-seq. a Comparison of human PBMC transcriptome profiling with different deep sequencing datasets. Wiggle tracks show an enrichment of 10x Chromium reads (rose) at the 3′ end of the GAPDH gene locus, close to its known PAS (GENCODE), while reads of TruSeq RNA-seq (gray) and Smart-seq2 (dark blue) cover the whole gene body. Data were retrieved from published PBMC TruSeq RNA-seq, Smart-seq2, and 10x Chromium . b Distribution of deep sequencing reads on mRNA genes. Pileup of deep sequencing reads from the same published datasets ( a ) indicates enrichment of 10x Chromium reads (rose) at 3′ ends of genes, compared to coverage of gene bodies by TruSeq RNA-seq (gray) and Smart-seq2 (dark blue) data. The distribution of PASs on mRNA genes were annotated in GENCODE. c Schematic of a stepwise SCAPTURE pipeline for single-cell PAS calling, filtering, transcript calculating, and APA analyzing. Top left, calling peaks from 3′ tag-based scRNA-seq (the first step). Top right, identifying high-confidence PASs with an embedded deep learning neural network DeepPASS (the second step). Bottom right, quantifying PASs to represent transcript expression at a single-cell resolution (the third step). Bottom left, applying SCAPTURE to APA analysis at single-cell level (the fourth step). See “ ” section for details

Article Snippet: Fig. 5 Identifying altered PAS usages upon SARS-CoV-2 infection. a Less proximal PAS usage in monocytes than other immune cell types, such as DCs, in six PBMC scRNA-seq datasets from 10x Genomics. b Schematic of APA analysis by SCAPTURE to compare PAS usage changes between healthy individuals and severe COVID-19 patients with SARS-CoV-2 infection at single-cell level.

Techniques: Comparison, Sequencing, RNA Sequencing, Expressing

Journal: Genome Biology

Article Title: SCAPTURE: a deep learning-embedded pipeline that captures polyadenylation information from 3′ tag-based RNA-seq of single cells

doi: 10.1186/s13059-021-02437-5

Figure Lengend Snippet: Applying SCAPTURE pipeline to study PASs in human PBMC datasets. a Application of SCAPTURE to identify exonic PASs from 3′ tag-based scRNA-seq datasets. Left, statistics of called peaks and identified PASs in six PBMC scRNA-seq datasets from 10x Genomics. Right, comparison of SCAPTURE-identified exonic PASs with known exonic PASs. Overlapped or non-overlapped exonic PASs identified by SCAPTURE were analyzed with known PAS annotation shown in Fig. a. b Nucleotide distribution of sequences at known (left), overlapped (middle), and non-overlapped (right) exonic PASs. Upstream (–) 100 bp to downstream (+) 100 bp sequences of PASs were analyzed. c Validation of SCAPTURE-identified non-overlapped PASs. Four such cases (salmon rectangle) in the NADK , PPP2R5C , MTX1 , and SMIM11B gene loci highlighted with their polyadenylation signals (AAUAAA motif) and further designed for validation by PCR as previously reported . A SCAPTURE-identified PAS that was previously annotated in the NADK gene locus (purple rectangle) was shown as a positive control. Top panel, TruSeq RNA-seq (gray), Smart-seq2 (dark blue), and 10x Chromium (rose) scRNA-seq from published datasets . Middle panel, wiggle tracks from one of 10x Genomics PBMC scRNA-seq datasets (the 10 k PBMC dataset) were indicated. Bottom panel, Ribo–RNA-seq datasets of bulk cells from normal (gray) and SLE patient (dark red) PBMC samples . PCR primers were shown in the bottom for experimental validation. See “ ” section for details. Asterisk, non-specific RNA-seq signals due to multiple mapping. d Validation of SCAPTURE-identified non-overlapped exonic PASs. Corresponding PCR products of these non-overlapped exonic PASs (salmon arrow) were shown with correct sizes. One overlapped exonic PAS (purple arrow) that was also identified by SCAPTURE in the NADK gene locus was shown as the positive control. As negative controls, no PCR products from downstream regions of these SCAPTURE-identified non-overlapped PASs were detected. Mixed PBMC RNA samples from SLE patients previously examined by Liu et al. were used for this validation. Of note, additional nested primers were designed to validate the SCAPTURE-identified PAS in the PPP2R5C locus. e Comparison of signature poly(A) signal motifs from exonic PASs identified by SCAPTURE or by two other recently reported pipelines, Sierra and scAPA . f Nucleotide distribution of sequences around exonic PASs identified by SCAPTURE, Sierra, or scAPA in six PBMC datasets from 10x Genomics. Upstream (–) 100 bp to downstream (+) 100 bp sequences of PASs were analyzed

Article Snippet: Fig. 5 Identifying altered PAS usages upon SARS-CoV-2 infection. a Less proximal PAS usage in monocytes than other immune cell types, such as DCs, in six PBMC scRNA-seq datasets from 10x Genomics. b Schematic of APA analysis by SCAPTURE to compare PAS usage changes between healthy individuals and severe COVID-19 patients with SARS-CoV-2 infection at single-cell level.

Techniques: Comparison, Biomarker Discovery, Positive Control, RNA Sequencing

Journal: Genome Biology

Article Title: SCAPTURE: a deep learning-embedded pipeline that captures polyadenylation information from 3′ tag-based RNA-seq of single cells

doi: 10.1186/s13059-021-02437-5

Figure Lengend Snippet: Using quantified PASs to represent distinct transcript expression for refining cell identities. a Schematic of single-cell analyses with either differential gene expression (DGE) or differential transcript expression (DTE) with SCAPTURE-identified and SCAPTURE-quantified PASs. b Numbers of expressed transcripts per gene in combined six PBMC datasets from 10x Genomics. c Number of expressed genes or transcripts in combined six PBMC datasets from 10x Genomics. d Gene distribution of top 2000 highly variable PAS-based transcript features. Of note, the majority of these top variable features are associated with genes expressing APA transcripts. e Phylogenetic analysis of cell types clustered by DTE with SCAPTURE (left) or by DGE with the conventional Seurat protocol (right). f Cross comparison of cell types assigned by DTE or DGE. g Comparison of marker gene expression between DGE-assigned major cell types (black) and DGE-unassigned but DTE-assigned cells (red). Of note, among the DGE-unassigned but DTE-assigned cells, some showed similar marker gene expression patterns as natural killer cells (NKCs; 68 out of 237) or dendritic cells (DCs; 48 out of 237). h Phylogenetic re-analysis of cell types clustered by DTE (left) or by DGE (right). DGE-unassigned but DTE-assigned cells are highlighted in both clusters (red)

Article Snippet: Fig. 5 Identifying altered PAS usages upon SARS-CoV-2 infection. a Less proximal PAS usage in monocytes than other immune cell types, such as DCs, in six PBMC scRNA-seq datasets from 10x Genomics. b Schematic of APA analysis by SCAPTURE to compare PAS usage changes between healthy individuals and severe COVID-19 patients with SARS-CoV-2 infection at single-cell level.

Techniques: Expressing, Refining, Gene Expression, Comparison, Marker

Journal: Genome Biology

Article Title: SCAPTURE: a deep learning-embedded pipeline that captures polyadenylation information from 3′ tag-based RNA-seq of single cells

doi: 10.1186/s13059-021-02437-5

Figure Lengend Snippet: Identifying altered PAS usages upon SARS-CoV-2 infection. a Less proximal PAS usage in monocytes than other immune cell types, such as DCs, in six PBMC scRNA-seq datasets from 10x Genomics. b Schematic of APA analysis by SCAPTURE to compare PAS usage changes between healthy individuals and severe COVID-19 patients with SARS-CoV-2 infection at single-cell level. See “ ” section for details. c UMAP plots showing DTE-based single-cell clustering of PBMC scRNA-seq datasets from healthy individuals and COVID-19 patients by SCAPTURE. Left, single cells labeled with disease group. Right, single cells labeled with annotated cell type. d UMAP plots showing global preference of proximal PAS usage in healthy individuals and COVID-19 patients at single-cell level. “ ” section for details. e Proximal PAS usage in immune cell types from PBMC scRNA-seq datasets of healthy individuals (left) and COVID-19 (right) patients. f Genes with altered PAS usage between monocytes and DCs. Examples of known genes with reported functions, such as activation of innate/adaptive immune response, antigen processing/presentation via MHC, or dendritic spine development, were shown. g Global preference of proximal PAS usage in different immune cell types from COVID-19 samples. Statistical significance was assessed by Kolmogorov-Smirnov test. h Preferential proximal PAS usage of some immunoglobulin genes in B cells (left) and plasma cells (right) of COVID-19 samples. Statistical significance was assessed with Seurat (“ ” section)

Article Snippet: Fig. 5 Identifying altered PAS usages upon SARS-CoV-2 infection. a Less proximal PAS usage in monocytes than other immune cell types, such as DCs, in six PBMC scRNA-seq datasets from 10x Genomics. b Schematic of APA analysis by SCAPTURE to compare PAS usage changes between healthy individuals and severe COVID-19 patients with SARS-CoV-2 infection at single-cell level.

Techniques: Infection, Labeling, Activation Assay, Clinical Proteomics

Journal: BMC Bioinformatics

Article Title: CAraCAl: CAMML with the integration of chromatin accessibility

doi: 10.1186/s12859-024-05833-3

Figure Lengend Snippet: CAMML scores for the built-in NK cell gene set on the joint scRNA-seq and scATAC-seq PBMC data from 10X Genomics. A UMAP of the joint scRNA-seq/scATAC-seq data, colored by identity. B The CAMML scores for NK cells on the scRNA-seq assay from this dataset. C The CAMML scores for NK cells on the gene activity estimates generated from the scATAC-seq data. D The integrated CAMML scores generated on scRNA-seq and scATAC-seq data . For clarity of the comparison, the UMAP computed on the scATAC-seq based gene activity estimates is used to illustrate the scores in all three plots

Article Snippet: Three publicly available datasets were used to evaluate the performance of CAraCAl: joint scRNA-seq/scATAC-seq peripheral blood mononuclear cell (PBMC) data from 10X Genomics [ ], scATAC-seq data from bone marrow and PBMCs that was manually annotated for cell identity [ ], and scATAC-seq data for sorted PBMC immune cell types [ ].

Techniques: Activity Assay, Generated, Comparison

Journal: Cell reports methods

Article Title: Single-cell signatures identify microenvironment factors in tumors associated with patient outcomes.

doi: 10.1016/j.crmeth.2024.100799

Figure Lengend Snippet: Figure 1. scBeacon workflow and validation (A) scBeacon workflow. Individual scRNA-seq datasets are clustered using Louvain clustering. Cluster centroids are ranked and then compared to each other using a reciprocal top-K enrichment (RTKE) metric. High-scoring cluster pairs that exceed an empirically determined threshold are retained as a graph for further clustering to identify exemplars from associated groups of meta-clusters. Exemplar centroids are computed by averaging the cluster-ranked centroids and recorded as exemplar signatures, assumed to be proxies of cell-type signatures for downstream analysis (see STAR Methods). (B‒D) tSNSE plots of PBMC scRNA-seq centroids using different transformations of the count-based data or similarity calculations between centroids. Left panel in each plot shows cells colored by single-cell sequencing technology platform (10x version 1 or 2 chemistries, green; 10x version 3 chemistry, aqua; 10x Chromium version 3 chemistry, light green; 10x Chromium version 2A chemistry, red; 10x Chromium version 2B chemistry, orange; CEL-Seq2, light blue; Drop- seq, medium blue; inDrops, dark blue; Seq-Well, purple; Smart-seq2, pink). Right panels in each plot show cells colored by cell type (T cells, light blue; B cells, red; monocytes, green). (B) Centroids represent vectors of count-based data (transcripts per million reads, TPMs). (C) Same as (B) but centroids were rank normalized. (D) Same as (C) but using the matrix of RTKE similarity metrics as input to tSNE.

Article Snippet: REAGENT or RESOURCE SOURCE IDENTIFIER Deposited data Single Cell Expression Atlas (SCEA) scRNA-seq data Papatheodorou et al.8 see Table S1 TCGA bulk RNA-seq data Hoadley et al.21 see Table S1 scRNA-Seq data from a human head and neck cancer dataset Puram et al.51 GSE103322 human melanoma dataset Tirosh et al.52 GSE72056 MSigDB gene set collection Liberzon et al.53 https://www.gsea-msigdb.org/gsea/msigdb/ scRNA-Seq PBMC datasets Ding et al.54 see Table S1 scRNA-Seq PBMC 10X genomics https://www.10xgenomics.com/datasets/ 10-k-pbm-cs-from-a-healthy-donorv-3-chemistry-3-standard-3-0-0 Bulk RNA-Seq data for each of melanoma cell lines Pawlikowski et al.55 see Table S1 Software and algorithms CIBERSORT Chen et al.56 https://cibersortx.stanford.edu/ Tumor Cell Type (TCT) UCSC TumorMap Newton et al.20 bit.ly/TCTmap_217exemplars Seurat Stuart et al.57 https://satijalab.org/seurat/ PanglaoDB Savi c et al.58 https://panglaodb.se/ Harmonizome Rouillard et al.59 https://maayanlab.cloud/Harmonizome/ Original code for data analysis This paper https://doi.org/10.6084/m9.figshare.25814632.v1

Techniques: Biomarker Discovery, Sequencing

Journal: Cell reports methods

Article Title: Single-cell signatures identify microenvironment factors in tumors associated with patient outcomes.

doi: 10.1016/j.crmeth.2024.100799

Figure Lengend Snippet: Figure 2. Validation of scBeacon workflow in synthetic mixtures of an scRNA-seq mel- anoma dataset (A) Correlation between the true mixture proportion of in silico mixtures from an scRNA-seq melanoma dataset to the deconvolution estimates of using a count-based signature matrix from a single PBMC scRNA-seq dataset (10X, v2). Red line marks the correct estimate (x = y). Cell-type ratios are normalized to sum up to 1. (B) Same as (A) but using a rank-normalized signature matrix from the combination of multiple PBMC scRNA-seq datasets: all PBMC datasets from Figures 1B–1D, except Smart-seq2: 10X chemistry v1–v3, CEL-Seq2, Drop-Seq, inDrops, and Seq-Well. (RMSE = root-mean-square error, corr = Pearson correlation).

Article Snippet: REAGENT or RESOURCE SOURCE IDENTIFIER Deposited data Single Cell Expression Atlas (SCEA) scRNA-seq data Papatheodorou et al.8 see Table S1 TCGA bulk RNA-seq data Hoadley et al.21 see Table S1 scRNA-Seq data from a human head and neck cancer dataset Puram et al.51 GSE103322 human melanoma dataset Tirosh et al.52 GSE72056 MSigDB gene set collection Liberzon et al.53 https://www.gsea-msigdb.org/gsea/msigdb/ scRNA-Seq PBMC datasets Ding et al.54 see Table S1 scRNA-Seq PBMC 10X genomics https://www.10xgenomics.com/datasets/ 10-k-pbm-cs-from-a-healthy-donorv-3-chemistry-3-standard-3-0-0 Bulk RNA-Seq data for each of melanoma cell lines Pawlikowski et al.55 see Table S1 Software and algorithms CIBERSORT Chen et al.56 https://cibersortx.stanford.edu/ Tumor Cell Type (TCT) UCSC TumorMap Newton et al.20 bit.ly/TCTmap_217exemplars Seurat Stuart et al.57 https://satijalab.org/seurat/ PanglaoDB Savi c et al.58 https://panglaodb.se/ Harmonizome Rouillard et al.59 https://maayanlab.cloud/Harmonizome/ Original code for data analysis This paper https://doi.org/10.6084/m9.figshare.25814632.v1

Techniques: Biomarker Discovery, In Silico