Journal: iScience

Article Title: A unique malignant cell type per patient tumor encoded in each cancer cell transcriptome

doi: 10.1016/j.isci.2026.115139

Figure Lengend Snippet: Tumor-driven malignant cell grouping is independent of technical bias or tumor-specific characteristics in gene expression for all cancers studied See also and . Bar plots illustrate Normalized Mutual Information (NMI) values obtained after data manipulation i.e., following either dropout imputation (orange outline), using only genes detected in all tumors (blue outline), using the 500 most highly variable genes (HVG, green outline), as well as after exclusion of the 500 most variable genes (yellow outline), or after binarization of gene expression (pink outline). Horizontal solid and dotted lines correspond to the NMI values obtained from analyses performed without data manipulation of malignant and, when available, non-malignant cells, respectively. (A) GB malignant cells from the scRNA-seq dataset of Darmanis et al. Bar plot (a1) and alluvial plot (a2) representations of the contribution of each tumor to each cluster following data manipulations. Compare with the alluvial plot of A. (B) GB malignant cells from independent datasets obtained with either SMART-seq2 (N-S), 10X Genomics single-cell RNA-seq (N-10X, PA-10X), or in-house sequencing technologies (Yuan, Yu). (C) Malignant cells from other types of brain tumors of the adult (OGD and IDH1 MUT Astrocytoma) and the infant (H3K27M glioma, diffuse gliomas, and pediatric gliomas). IDH1 MUT : isocitrate dehydrogenase 1, mutated form. OGD: IDH1 MUT 1p/19q-codeleted Oligodendroglioma. SMART-seq2 technology. (D) Malignant cells from non-brain cancers, including head and neck cancers (SMART-seq2 sequencing technology), primary breast cancers (Fluidigm C1 technology), melanoma (SMART-seq2 technology), and ovarian cancers (10X Genomics technology). (E) Tumor-driven cancer cell grouping consistently decreases only following the binarization of gene expression. NMI values are presented with respect to the NMI value obtained without data manipulation for malignant cell grouping of each dataset (normalized NMI). 20 independent datasets analyzed corresponding to the datasets illustrated in panels A to D and datasets of pancreatic adenocarcinoma, lung cancer, acute lymphoblastic leukemia, multiple myeloma, and Non-Hodgkin lymphoma, illustrated in . Dotted line marks the NMI value without data manipulation. Mean ± SD values are shown. One sample t test, p < 0.01. (F) Reduced NMI values of non-malignant cell grouping compared to NMI values of malignant cells observed using two other analytical approaches, the Leiden clustering method based on community detection and DCA for data reduction. Dot plot illustrates reduced NMI values of non-malignant cell grouping compared to NMI values of malignant cells from datasets illustrated in panels A to D. Mean ± SD values are shown. Mann-Whitney, ∗: p < 0.0001. M: malignant cells and N: non-malignant cells. See B for the corresponding bar plot.

Article Snippet: Compare with the alluvial plot of A. (B) GB malignant cells from independent datasets obtained with either SMART-seq2 (N-S), 10X Genomics single-cell RNA-seq (N-10X, PA-10X), or in-house sequencing technologies (Yuan, Yu). (C) Malignant cells from other types of brain tumors of the adult (OGD and IDH1 MUT Astrocytoma) and the infant (H3K27M glioma, diffuse gliomas, and pediatric gliomas).

Techniques: Gene Expression, Single Cell, RNA Sequencing, Sequencing, MANN-WHITNEY

Journal: Advanced Science

Article Title: EIF5A Couples Translational Control With Transcriptional Reprogramming Through Chromocenter Reorganization During Spermiogenesis

doi: 10.1002/advs.202517423

Figure Lengend Snippet: Increased chromatin accessibility in knockout round spermatids. A).Volcano plot of DEGs between control and Eif5a SKO samples.FDR<0.05. B.The average tag density plot(top pannel) and heatmaps (bottom pannel)around TSS (±3 kb) for the enrichment of ATAC‐seq reads in control and Eif5a SKO round spermatids. C).Plot shows Gain and Loss sites in all control and Eif5a SKO sample replicates. D).Venn diagram showing the overlap of differential peaks identified by CUT&Tag and ATAC‐seq between WT and CKO groups. E).Four‐quadrant scatter plot comparing the log 2 fold changes of significantly differential peaks (FDR < 0.05) from H3K4me3 CUT&Tag (x‐axis) and ATAC‐seq (y‐axis). The Pearson correlation coefficient for the compared data is 0.468. Pearson's *r* = 0.468. F) Venn diagram show shared genes between ATAC‐seq (FDR<0.05) and Smart‐seq2 (FDR<0.05, FC>1.5).

Article Snippet: The Smart‐seq2 library sequencing was performed by Novogene on Illumina platforms, generating 150 bp paired‐end reads.

Techniques: Knock-Out, Control

Journal: Advanced Science

Article Title: EIF5A Couples Translational Control With Transcriptional Reprogramming Through Chromocenter Reorganization During Spermiogenesis

doi: 10.1002/advs.202517423

Figure Lengend Snippet: Proteomic alterations associated with transcriptional changes induced by Eif5a deletion. A).Venn diagram of shared genes between Smart‐seq2 (P value<0.05, FC>1.5) and Proteomics (P<0.05, FC>1.5) analyses. B).GO enrichment analysis based on the 119 commonly upregulated genes. C).QRT‐PCR analysis of candidate genes that were consistently dysregulated in both the transcriptome and proteome of Eif5a SKO testes. Data are presented as mean ± SD from three independent biological replicates (n = 3). Statistical significance was determined using a two‐tailed, unpaired Student's t‐test (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). D). Western blots show SPATA1, SPACA3 and SPACA9 proteins in Eif5a SKO and control mice. β‐Actin served as the loading control. E). IGV visualization of genomic regions harboring acrosome‐related ( Spaca3, Ly6K,Spaca9,Spata1,Lamp2 ) and microtubule‐associated ( Ccdc169, Dynlt3 ) genes. Top: ATAC‐seq tracks showing chromatin accessibility in control (blue) versus SKO (red) round spermatids. Bottom: Corresponding Smart‐seq2 coverage.

Article Snippet: The Smart‐seq2 library sequencing was performed by Novogene on Illumina platforms, generating 150 bp paired‐end reads.

Techniques: Quantitative RT-PCR, Two Tailed Test, Western Blot, Control