American Society of Hematology

Figure 1.

$Figure 1. Either TGF-β signaling or SKI-correlated gene signatures identify activated TGF-β signaling and deregulated splicing in early-stage MDS. (A) Principal component analysis (PCA) plot for bulk RNA-Seq data for CD34+ bone marrow cells organized by a TGF-β signaling gene expression signature (23 healthy donors, 44 MDS patients). Each dot represents 1 patient sample, healthy donors are indicated by black borders, and the color represents strong correlation to SKI or SKIL. (B) Heatmap of the top SKI-correlated genes. Each column represents 1 sample and each row represents 1 gene. Top: gene expression clusters were generated in AltAnalyze with the key genes indicated on the right; bottom: clinical characterizations are noted. Mutations in splicing factors are noted (U2AF1-mut, SF3B1-mut, and SRSF2-mut) (supplemental Figure 1A). (C) Scatter plot displaying the false discovery rate (FDR) P value vs Z-score of individual biological processes enriched by the differentially (top) upregulated or (bottom) downregulated genes in SKI-high MDS samples compared with controls. Each dot represents a biological process, and the dotted line indicates an FDR P value of .05. (D) PCA plot from panel A, in which the color represents increasing expression of HNRNPK. (E) Alternative splicing (AS) example: representative sashimi plots of the CSF3R locus reveals MDS-specific premature termination of the open reading frame. (F) Splicing concordance analysis: scatter plot displaying splicing event concordance vs the number of AS events in common between SKI-high MDS samples and each displayed splicing data set (eg, HNRNPK-KD). GPCR, G protein-coupled receptors; RA, refractory anemia; RAEB, refractory anemia with excess blasts; RARS, refractory anemia with ringed sideroblasts; TCA, tricarboxylic acid.$

Either TGF-β signaling or SKI-correlated gene signatures identify activated TGF-β signaling and deregulated splicing in early-stage MDS. (A) Principal component analysis (PCA) plot for bulk RNA-Seq data for CD34⁺ bone marrow cells organized by a TGF-β signaling gene expression signature (23 healthy donors, 44 MDS patients). Each dot represents 1 patient sample, healthy donors are indicated by black borders, and the color represents strong correlation to SKI or SKIL. (B) Heatmap of the top SKI-correlated genes. Each column represents 1 sample and each row represents 1 gene. Top: gene expression clusters were generated in AltAnalyze with the key genes indicated on the right; bottom: clinical characterizations are noted. Mutations in splicing factors are noted (U2AF1-mut, SF3B1-mut, and SRSF2-mut) (supplemental Figure 1A). (C) Scatter plot displaying the false discovery rate (FDR) P value vs Z-score of individual biological processes enriched by the differentially (top) upregulated or (bottom) downregulated genes in SKI-high MDS samples compared with controls. Each dot represents a biological process, and the dotted line indicates an FDR P value of .05. (D) PCA plot from panel A, in which the color represents increasing expression of HNRNPK. (E) Alternative splicing (AS) example: representative sashimi plots of the CSF3R locus reveals MDS-specific premature termination of the open reading frame. (F) Splicing concordance analysis: scatter plot displaying splicing event concordance vs the number of AS events in common between SKI-high MDS samples and each displayed splicing data set (eg, HNRNPK-KD). GPCR, G protein-coupled receptors; RA, refractory anemia; RAEB, refractory anemia with excess blasts; RARS, refractory anemia with ringed sideroblasts; TCA, tricarboxylic acid.

This Feature Is Available To Subscribers Only