Search

Additional file 1. Supplementary methods, Supplementary figures (Fig. S1-S9), and Supplementary tables (Table S1-S25).

Additional file 3 This pdf file contains two tables which include results on two related tasks (table 1, 2).

Additional file 1. This file contains the R markdown output for the analysis. The plots and code in this file were directly used to generate the statistics and figures in the manuscript.

Additional file 1: Figure S1. An example framework of the AECOX model with four hidden layers. Table S1. The network design of AECOX. Table S2. The hyper-parameters of AECOX to be searched. Table S3. Performances of testing set in TCGA Kidney Renal Clear Cell Carcinoma (KIRC) dataset. Bolded texts indicated optimal results among all models. Table S4. Individual model correlations (Pearson ρ) of mean TMB (Fig. 2). Figure S2. Relationship between concordance index and median TMB. Pearson ρ = − 0.30 (p-value = 7.75E-02). Figure S3. Relationship between concordance index and max TMB. Pearson ρ = − 0.40 (p-value = 1.68E-02). Figure S4. Relationship between concordance index and 20% tail TMB. Pearson ρ = − 0.32 (p-value = 5.51E-02). Figure S5. Relationship between concordance index and 10% tail TMB. Pearson ρ = − 0.32 (p-value = 5.93E-02). Figure S6. Relationship between concordance index and 5% tail TMB. Pearson ρ = − 0.30 (p-value = 7.45E-02). Table S5. Fine-tuned hyper-parameters of Cox-nnet (L2 penalty weight λ) across 12 cancer types and 5 experiments (folds). Table S6. Fine-tuned hyper-parameters of DeepSurv across 12 cancer types and 5 experiments (folds). Table S7. Fine-tuned hyper-parameters of AECOX across 12 cancer types and 5 experiments (folds). Note that we fixed λ2 = 0 to only impose L2 sparsity. Table S8. Fine-tuned hyper-parameters of Random Survival Forest (RSF) (number of the trees) across 12 cancer types and 5 experiments (folds). Table S9. Fine-tuned hyper-parameters of SVM (α, weight of penalizing the squared hinge loss in the objective function) across 12 cancer types and 5 experiments (folds). Table S10. Model-wised performances comparison at pan-cancer level (12 TCGA (The Cancer Genome Atlas) cancer types) by pairwise paired t-test, according to metrics concordance index and p-value of log-rank test. Note that for concordance index, larger t-statistic/coefficient indicated better performance at pan-cancer level, while the p-value of log-rank test was on the contrary. Table S11. Model-wised performances comparison at pan-cancer level (12 TCGA (The Cancer Genome Atlas) cancer types) by linear mixed-effects models test, according to metrics concordance index and p-value of log-rank test. Note that for concordance index, larger t-statistic/coefficient indicated better performance at pan-cancer level, while the p-value of log-rank test was on the contrary.

Diabetes affects >10% of adults worldwide and is caused by impaired production or response to insulin, resulting in chronic hyperglycemia. Pancreatic islet β-cells are the sole source of endogenous insulin and our understanding of β-cell dysfunction and death in type 2 diabetes (T2D) is incomplete. Single-cell RNA-seq data supports heterogeneity as an important factor in β-cell function and survival. However, it is difficult to identify which β-cell phenotypes are critical for T2D etiology and progression. Our goal was to prioritize specific disease-related β-cell subpopulations to better understand T2D pathogenesis and identify relevant genes for targeted therapeutics. To address this, we applied a deep transfer learning tool, DEGAS, which maps disease associations onto single-cell RNA-seq data from bulk expression data. Independent runs of DEGAS using T2D or obesity status identified distinct β-cell subpopulations. A singular cluster of T2D-associated β-cells was identified; however, β-cells with high obese-DEGAS scores contained two subpopulations derived largely from either non-diabetic or T2D donors. The obesity-associated non-diabetic cells were enriched for translation and unfolded protein response genes compared to T2D cells. We selected DLK1 for validation by immunostaining in human pancreas sections from healthy and T2D donors. DLK1 was heterogeneously expressed among β-cells and appeared depleted from T2D islets. In conclusion, DEGAS has the potential to advance our holistic understanding of the β-cell transcriptomic phenotypes, including features that distinguish β-cells in obese non-diabetic or lean T2D states. Future work will expand this approach to additional human islet omics datasets to reveal the complex multicellular interactions driving T2D.

Multiple Myeloma is an incurable plasma cell malignancy with a poor survival rate that is usually treated with immunomodulatory drugs (iMiDs) and proteosome inhibitors (PIs). The malignant plasma cells quickly become resistant to these agents causing relapse and uncontrolled growth of resistant clones. From whole genome sequencing (WGS) and RNA sequencing (RNA-seq) studies, different high-risk translocation, copy number, mutational, and transcriptional markers have been identified. One of these markers, PHF19 , epigenetically regulates cell cycle and other processes and has already been studied using RNA-seq. In this study a massive (325,025 cells and 49 patients) single cell multiomic dataset was generated with jointly quantified ATAC- and RNA-seq for each cell and matched genomic profiles for each patient. We identified an association between one plasma cell subtype with myeloma progression that we have called relapsed/refractory plasma cells (RRPCs). These cells are associated with 1q alterations, TP53 mutations, and higher expression of PHF19 . We also identified downstream regulation of cell cycle inhibitors in these cells, possible regulation of the transcription factor (TF) PBX1 on 1q, and determined that PHF19 may be acting primarily through this subset of cells., Competing Interests: CONFLICT OF INTEREST DISCLOSURE V.C., C.D., and M.N. are Genetech/Roche employees and hold Roche stock options.