Robust estimation of cancer and immune cell-type proportions from bulk tumor ATAC-Seq data.

Assay for Transposase-Accessible Chromatin sequencing (ATAC-Seq) is a widely used technique to explore gene regulatory mechanisms. For most ATAC-Seq data from healthy and diseased tissues such as tumors, chromatin accessibility measurement represents a mixed signal from multiple cell types. In this work, we derive reliable chromatin accessibility marker peaks and reference profiles for most non-malignant cell types frequently observed in the microenvironment of human tumors. We then integrate these data into the EPIC deconvolution framework (Racle et al., 2017) to quantify cell-type heterogeneity in bulk ATAC-Seq data. Our EPIC-ATAC tool accurately predicts non-malignant and malignant cell fractions in tumor samples. When applied to a human breast cancer cohort, EPIC-ATAC accurately infers the immune contexture of the main breast cancer subtypes.

The tea cultivar 'Chungui' with jasmine-like aroma: From genome and epigenome to quality.

'Chungui' is a newly promoted tea cultivar in China, renowned for producing oolong tea with a distinctive jasmine-like aroma. However, the genetic basis of this unique aroma remains unclear. In this study, the 'Chungui' genome, one of the most complete and well-annotated tea genomes, was assembled using PacBio HiFi reads and Hi-C sequencing. Through comparative analysis with typical jasmine flower volatiles, eight core compounds responsible for this aroma were identified. Further research revealed that the jasmine-like aroma in 'Chungui' is regulated by a coordinated mechanism involving a significant increase in chromatin accessibility and the demethylation of CHH and CHG in the promoter regions of key aroma-related genes during oolong tea processing. The study proposes that the formation of this unique aroma is driven by the synergistic effect of enhanced chromatin accessibility and reduced methylation, which together lead to the robust upregulation of genes involved in the biosynthesis of these core aroma components. These results provide a molecular foundation for understanding the unique jasmine-like aroma of 'Chungui' tea and sets the stage for future studies to explore the roles of these regulatory mechanisms in aroma formation.

Dynamic chromatin accessibility and transcriptome changes following PDGF-BB treatment of bone-marrow derived mesenchymal stem cells.

BACKGROUND: Mesenchymal stem cells (MSCs) are multipotent stem cells that are under investigation for use in clinical trials because they are capable of self-renewal and differentiating into different cell types under defined conditions. Nonetheless, the therapeutic effects of MSCs have been constrained by low engraftment rates, cell fusion, and cell survival. Various strategies have been explored to improve the therapeutic efficacy of MSCs, with platelet-derived growth factor (PDGF)-BB emerging as a promising candidate. To enhance our comprehension of the impact of PDGF-BB on the gene expression profile and chromosomal accessibility of MSCs, RNA-sequencing and analysis of chromatin accessibility profiles were conducted on three human primary MSCs in culture, both with and without stimulation by PDGF-BB. RESULTS: Integrative analysis of gene expression and chromatin accessibility demonstrated that PDGF-BB treatment modified the chromatin accessibility landscape, marking regions for activation or repression through the AP-1 family transcription factors TEAD, CEBP, and RUNX2. These changes in AP-1 transcription factor expression, in turn, led to cell proliferation and differentiation potential towards osteoblasts, adipocytes, or chondrocytes. The degree of MSC differentiation varies among cells isolated from different donors. The presence of an enrichment of exosome-related genes is also noted among all the differentially expressed genes. CONCLUSIONS: In conclusion, the observed changes in AP-1 transcription factor expression not only induced cellular proliferation and differentiation, but also revealed variations in the degree of MSC differentiation based on donor-specific differences. Moreover, the enrichment of exosome-related genes among differentially expressed genes suggests a potential significant role for PDGF-BB in facilitating intercellular communication.

CYP1B1 promotes PARPi-resistance via histone H1.4 interaction and increased chromatin accessibility in ovarian cancer.

INTRODUCTION: Ovarian cancer is the most lethal gynecological cancer and presents significant therapeutic challenges. The discovery of synthetic lethality between PARP inhibitors (PARPi) and homologous recombination deficiency marked a new era in treating BRCA1/2-mutated tumors. However, PARPi resistance remains a major clinical challenge. METHODS: RNA sequencing was used to identify genes altered by PARPi treatment and LC-MS was used to detect proteins interacting with CYP1B1. Resistance mechanisms were explored through ATAC-seq and gene expression manipulation. Additional techniques, including micrococcal nuclease digestion assays, DAPI staining, and fluorescence microscopy, were used to assess changes in nuclear morphology and chromatin accessibility. RESULTS: The gradual exposure of Olaparib has developed a PARPi-resistant cell line, A2780-OlaR, which exhibits significant upregulation of CYP1B1 at both RNA and protein levels. Down-regulating CYP1B1 expression or using specific inhibitors decreased the cellular response to Olaparib. Linker histone H1.4 was identified as associated with CYP1B1. ATAC-seq showed differential chromatin accessibility between A2780-OlaR and parental cells, indicating that the downregulation of H1.4 was associated with increased chromatin accessibility and higher cell viability after Olaparib treatment. CONCLUSION: Our findings reveal a novel role for CYP1B1 in driving PARPi resistance through distinct molecular mechanisms in A2780-OlaR. This study highlights the importance of chromatin accessibility in PARPi efficacy and suggests the CYP1B1/H1.4 axis as a promising therapeutic target for overcoming drug resistance in ovarian cancer, offering potentially therapeutic benefits.

Definition of regulatory elements and transcription factors controlling porcine immune cell gene expression at single cell resolution using single nucleus ATAC-seq.

The transcriptome of porcine peripheral blood mononuclear cells (PBMC) at single cell (sc) resolution is well described, but little is understood about the cis-regulatory mechanism behind scPBMC gene expression. Here, we profiled the open chromatin landscape of porcine PBMC that define cis-regulatory elements and mechanism contributing to the transcription using single nucleus ATAC sequencing (snATAC-seq). Approximately 22 % of the identified peaks overlapped with annotated transcription start sites (TSS). Using clustering based on open chromatin pattern similarity, we demonstrate that cell type annotations using snATAC-seq are highly concordant to that reported for sc RNA sequencing (scRNA-seq). The differentially accessible peaks (DAPs) for each cell type were characterized and the pattern of accessibility of the DAPs near cell type markers across cell types was similar to that of the average gene expression level of corresponding marker genes. Additionally, we found that peaks identified in snATAC-seq have the potential power to predict the cell type specific transcription starting site (TSS). We identified both transcription factors (TFs) whose binding motif were enriched in cell type DAPs of multiple cell types and cell type specific TFs by conducting transcription factor binding motif (TFBM) analysis. Furthermore, we identified the putative enhancer or promoter regions bound by TFs for each differentially expressed gene (DEG) with a DAP that overlapped with its TSS by generating cis-co-accessibility networks (CCAN). To predict the regulators of such DEGs, TFBM analysis was performed for each CCAN. The regulator TF-target DEG pairs predicted in this way were largely consistent with the results reported in the ENCODE Transcription Factor Targets Dataset (TFTD). This snATAC-seq approach provides insights into the regulation of chromatin accessibility landscape of porcine PBMCs and enables discovery of TFs predicted to control DEG through binding regulatory elements whose chromatin accessibility correlates with the DEG promoter region.

Genome-wide epigenetic dynamics of tea leaves under mechanical wounding stress during oolong tea postharvest processing.

Understanding the epigenetic responses to mechanical wounding stress during the postharvest processing of oolong tea provides insight into the reprogramming of the tea genome and its impact on tea quality. Here, we characterized the 5mC DNA methylation and chromatin accessibility landscapes of tea leaves subjected to mechanical wounding stress during the postharvest processing of oolong tea. Analysis of the differentially methylated regions and preferentially accessible promoters revealed many overrepresented TF-binding motifs, highlighting sets of TFs that are likely important for the quality of oolong tea. Within these sets, we constructed a chromatin accessibility-mediated gene regulatory network specific to mechanical wounding stress. In combination with the results of the TF-centred yeast one-hybrid assay, we identified potential binding sites of CsMYC2 and constructed a gene regulatory network centred on CsMYC2, clarifying the potential regulatory role of CsMYC2 in the postharvest processing of oolong tea. Interestingly, highly accessible chromatin and hypomethylated cytosine were found to coexist in the promoter region of the indole biosynthesis gene (tryptophan synthase ß-subunit, CsTSB) under wounding stress, which indicates that these two important epigenetic regulatory mechanisms are jointly involved in regulating the synthesis of indole during the postharvest processing of oolong tea. These findings improve our understanding of the epigenetic regulatory mechanisms involved in quality formation during the postharvest processing of oolong tea.

Enhlink infers distal and context-specific enhancer-promoter linkages.

Enhlink is a computational tool for scATAC-seq data analysis, facilitating precise interrogation of enhancer function at the single-cell level. It employs an ensemble approach incorporating technical and biological covariates to infer condition-specific regulatory DNA linkages. Enhlink can integrate multi-omic data for enhanced specificity, when available. Evaluation with simulated and real data, including multi-omic datasets from the mouse striatum and novel promoter capture Hi-C data, demonstrate that Enhlink outperfoms alternative methods. Coupled with eQTL analysis, it identified a putative super-enhancer in striatal neurons. Overall, Enhlink offers accuracy, power, and potential for revealing novel biological insights in gene regulation.

Temporally discordant chromatin accessibility and DNA demethylation define short and long-term enhancer regulation during cell fate specification.

Epigenetic mechanisms govern the transcriptional activity of lineage-specifying enhancers; but recent work challenges the dogma that joint chromatin accessibility and DNA demethylation are prerequisites for transcription. To understand this paradox, we established a highly-resolved timeline of DNA demethylation, chromatin accessibility, and transcription factor occupancy during neural progenitor cell differentiation. We show thousands of enhancers undergo rapid, transient accessibility changes associated with distinct periods of transcription factor expression. However, most DNA methylation changes are unidirectional and delayed relative to chromatin dynamics, creating transiently discordant epigenetic states. Genome-wide detection of 5-hydroxymethylcytosine further revealed active demethylation begins ahead of chromatin and transcription factor activity, while enhancer hypomethylation persists long after these activities have dissipated. We demonstrate that these timepoint specific methylation states predict past, present and future chromatin accessibility using machine learning models. Thus, chromatin and DNA methylation collaborate on different timescales to mediate short and long-term enhancer regulation during cell fate specification.

Multiome Perturb-seq unlocks scalable discovery of integrated perturbation effects on the transcriptome and epigenome.

Single-cell CRISPR screens link genetic perturbations to transcriptional states, but high-throughput methods connecting these induced changes to their regulatory foundations are limited. Here we introduce Multiome Perturb-seq, extending single-cell CRISPR screens to simultaneously measure perturbation-induced changes in gene expression and chromatin accessibility. We apply Multiome Perturb-seq in a CRISPRi screen of 13 chromatin remodelers in human RPE-1 cells, achieving efficient assignment of sgRNA identities to single nuclei via an improved method for capturing barcode transcripts from nuclear RNA. We organize expression and accessibility measurements into coherent programs describing the integrated effects of perturbations on cell state, finding that ARID1A and SUZ12 knockdowns induce programs enriched for developmental features. Pseudotime analysis of perturbations connects accessibility changes to changes in gene expression, highlighting the value of multimodal profiling. Overall, our method provides a scalable and simply implemented system to dissect the regulatory logic underpinning cell state.

Single-cell multi-modal chromatin profiles revealing epigenetic regulations of cells in hepatocellular carcinoma.

BACKGROUND: Various epigenetic regulations systematically govern gene expression in cells involving various biological processes. Dysregulation of the epigenome leads to aberrant transcriptional programs and subsequently results in diseases, such as cancer. Therefore, comprehensive profiling epigenomics is essential for exploring the mechanisms underlying gene expression regulation during development and disease. METHODS: In this study, we developed single-cell chromatin proteins and accessibility tagmentation (scCPA-Tag), a multi-modal single-cell epigenetic profile capturing technique based on barcoded Tn5 transposases and a droplet microfluidics platform. scCPA-Tag enables the simultaneous capture of DNA profiles of histone modification and chromatin accessibility in the same cell. RESULTS: By applying scCPA-Tag to K562 cells and a hepatocellular carcinoma (HCC) sample, we found that the silence of several chromatin-accessible genes can be attributed to lysine-27-trimethylation of the histone H3 tail (H3K27me3) modification. We characterized the epigenetic features of the tumour cells and different immune cell types in the HCC tumour tissue by scCPA-Tag. Besides, a tumour cell subtype (C2) with more aggressive features was identified and characterized by high chromatin accessibility and a lower abundance of H3K27me3 on tumour-promoting genes. CONCLUSIONS: Our multi-modal scCPA-Tag provides a comprehensive approach for exploring the epigenetic landscapes of heterogeneous cell types and revealing the mechanisms of gene expression regulation during developmental and pathological processes at the single-cell level. HIGHLIGHTS: scCPA-Tag offers a highly efficient and high throughput technique to simultaneously profile histone modification and chromatin accessibility within a single cell. scCPA-Tag enables to uncover multiple epigenetic modification features of cellular compositions within tumor tissues. scCPA-Tag facilitates the exploration of the epigenetic landscapes of heterogeneous cell types and provides the mechanisms governing gene expression regulation.

ΔNp63 bookmarks and creates an accessible epigenetic environment for TGFß-induced cancer cell stemness and invasiveness.

BACKGROUND: p63 is a transcription factor with intrinsic pioneer factor activity and pleiotropic functions. Transforming growth factor ß (TGFß) signaling via activation and cooperative action of canonical, SMAD, and non-canonical, MAP-kinase (MAPK) pathways, elicits both anti- and pro-tumorigenic properties, including cell stemness and invasiveness. TGFß activates the ΔNp63 transcriptional program in cancer cells; however, the link between TGFß and p63 in unmasking the epigenetic landscape during tumor progression allowing chromatin accessibility and gene transcription, is not yet reported. METHODS: Small molecule inhibitors, including protein kinase inhibitors and RNA-silencing, provided loss of function analyses. Sphere formation assays in cancer cells, chromatin immunoprecipitation and mRNA expression assays were utilized in order to gain mechanistic evidence. Mass spectrometry analysis coupled to co-immunoprecipitation assays revealed novel p63 interactors and their involvement in p63-dependent transcription. RESULTS: The sphere-forming capacity of breast cancer cells was enhanced upon TGFß stimulation and significantly decreased upon ΔNp63 depletion. Activation of TGFß signaling via p38 MAPK signaling induced ΔNp63 phosphorylation at Ser 66/68 resulting in stabilized ΔNp63 protein with enhanced DNA binding properties. TGFß stimulation altered the ratio of H3K27ac and H3K27me3 histone modification marks, pointing towards higher H3K27ac and increased p300 acetyltransferase recruitment to chromatin. By silencing the expression of ΔNp63, the TGFß effect on chromatin remodeling was abrogated. Inhibition of H3K27me3, revealed the important role of TGFß as the upstream signal for guiding ΔNp63 to the TGFß/SMAD gene loci, as well as the indispensable role of ΔNp63 in recruiting histone modifying enzymes, such as p300, to these genomic regions, regulating chromatin accessibility and gene transcription. Mechanistically, TGFß through SMAD activation induced dissociation of ΔNp63 from NURD or NCOR/SMRT histone deacetylation complexes, while promoted the assembly of ΔNp63-p300 complexes, affecting the levels of histone acetylation and the outcome of ΔNp63-dependent transcription. CONCLUSIONS: ΔNp63, phosphorylated and recruited by TGFß to the TGFß/SMAD/ΔNp63 gene loci, promotes chromatin accessibility and transcription of target genes related to stemness and cell invasion.

Establishment and Maintenance of Heat-Stress Memory in Plants.

Among the rich repertoire of strategies that allow plants to adapt to high-temperature stress is heat-stress memory. The mechanisms underlying the establishment and maintenance of heat-stress memory are poorly understood, although the chromatin opening state appears to be an important structural basis for maintaining heat-stress memory. The chromatin opening state is influenced by epigenetic modifications, making DNA and histone modifications important entry points for understanding heat-shock memory. Current research suggests that traditional heat-stress signaling pathway components might be involved in chromatin opening, thereby promoting the establishment of heat-stress memory in plants. In this review, we discuss the relationship between chromatin structure-based maintenance and the establishment of heat-stress memory. We also discuss the association between traditional heat-stress signals and epigenetic modifications. Finally, we discuss potential research ideas for exploring plant adaptation to high-temperature stress in the future.

Methyl-CpG-binding 2 K271 lactylation-mediated M2 macrophage polarization inhibits atherosclerosis.

Rationale: Posttranslational modifications of proteins have not been addressed in studies aimed at elucidating the cardioprotective effect of exercise in atherosclerotic cardiovascular disease (ASCVD). In this study, we reveal a novel mechanism by which exercise ameliorates atherosclerosis via lactylation. Methods: Using ApoE-/- mice in an exercise model, proteomics analysis was used to identify exercise-induced specific lactylation of MeCP2 at lysine 271 (K271). Mutation of the MeCP2 K271 lactylation site in aortic plaque macrophages was achieved by recombinant adenoviral transfection. Explore the molecular mechanisms by which motility drives MeCP2 K271 lactylation to improve plaque stability using ATAC-Seq, CUT &Tag and molecular biology. Validation of the potential target RUNX1 for exercise therapy using Ro5-3335 pharmacological inhibition. Results: we showed that in ApoE-/- mice, methyl-CpG-binding protein 2 (MeCP2) K271 lactylation was observed in aortic root plaque macrophages, promoting pro-repair M2 macrophage polarization, reducing the plaque area, shrinking necrotic cores, reducing plaque lipid deposition, and increasing collagen content. Adenoviral transfection, by introducing a mutant at lysine 271, overexpressed MeCP2 K271 lactylation, which enhanced exercise-induced M2 macrophage polarization and increased plaque stability. Mechanistically, the exercise-induced atheroprotective effect requires an interaction between MeCP2 K271 lactylation and H3K36me3, leading to increased chromatin accessibility and transcriptional repression of RUNX1. In addition, the pharmacological inhibition of the transcription factor RUNX1 exerts atheroprotective effects by promoting the polarization of plaque macrophages towards the pro-repair M2 phenotype. Conclusions: These findings reveal a novel mechanism by which exercise ameliorates atherosclerosis via MeCP2 K271 lactylation-H3K36me3/RUNX1. Interventions that enhance MeCP2 K271 lactylation have been shown to increase pro-repair M2 macrophage infiltration, thereby promoting plaque stabilization and reducing the risk of atherosclerotic cardiovascular disease. We also established RUNX1 as a potential drug target for exercise therapy, thereby providing guidance for the discovery of new targets.

Current genomic deep learning models display decreased performance in cell type-specific accessible regions.

BACKGROUND: A number of deep learning models have been developed to predict epigenetic features such as chromatin accessibility from DNA sequence. Model evaluations commonly report performance genome-wide; however, cis regulatory elements (CREs), which play critical roles in gene regulation, make up only a small fraction of the genome. Furthermore, cell type-specific CREs contain a large proportion of complex disease heritability. RESULTS: We evaluate genomic deep learning models in chromatin accessibility regions with varying degrees of cell type specificity. We assess two modeling directions in the field: general purpose models trained across thousands of outputs (cell types and epigenetic marks) and models tailored to specific tissues and tasks. We find that the accuracy of genomic deep learning models, including two state-of-the-art general purpose models-Enformer and Sei-varies across the genome and is reduced in cell type-specific accessible regions. Using accessibility models trained on cell types from specific tissues, we find that increasing model capacity to learn cell type-specific regulatory syntax-through single-task learning or high capacity multi-task models-can improve performance in cell type-specific accessible regions. We also observe that improving reference sequence predictions does not consistently improve variant effect predictions, indicating that novel strategies are needed to improve performance on variants. CONCLUSIONS: Our results provide a new perspective on the performance of genomic deep learning models, showing that performance varies across the genome and is particularly reduced in cell type-specific accessible regions. We also identify strategies to maximize performance in cell type-specific accessible regions.

Evidence of innate training in bovine γδ T cells following subcutaneous BCG administration.

The Bacillus Calmette Guerin (BCG) vaccine has been shown to induce non-specific protection against diseases other than tuberculosis in vaccinated individuals, attributed to the induction of trained immunity. We have previously demonstrated that BCG administration induces innate immune training in mixed peripheral blood mononuclear cells and monocytes in calves. Gamma Delta (γδ) T cells are non-conventional T cells that exhibit innate and adaptive immune system features. They are in higher proportion in the peripheral blood of cattle than humans or rodents and play an essential role in bovine immune response to pathogens. In the current study, we determined if BCG administration induced innate immune training in bovine γδ T cells. A group of 16 pre-weaned Holstein calves (2-4 d age) were enrolled in the study and randomly assigned to vaccine and control groups (n=8/group). The vaccine group received two doses of 106 colony forming units (CFU) BCG Danish strain subcutaneously, separated by 2 weeks. The control group remained unvaccinated. Gamma delta T cells were purified from peripheral blood using magnetic cell sorting three weeks after receiving the 1st BCG dose. We observed functional changes in the γδ T cells from BCG-treated calves shown by increased IL-6 and TNF-α cytokine production in response to in vitro stimulation with Escherichia coli LPS and PAM3CSK4. ATAC-Seq analysis of 78,278 regions of open chromatin (peaks) revealed that γδ T cells from BCG-treated calves had an altered epigenetic status compared to cells from the control calves. Differentially accessible peaks (DAP) found near the promoters of innate immunity-related genes like Siglec14, Irf4, Ifna2, Lrrfip1, and Tnfrsf10d were 1 to 4-fold more accessible in cells from BCG-treated calves. MOTIF enrichment analysis of the sequences within DAPs, which explores transcription factor binding motifs (TFBM) upstream of regulatory elements, revealed TFBM for Eomes and IRF-5 were among the most enriched transcription factors. GO enrichment analysis of genes proximal to the DAPs showed enrichment of pathways such as regulation of IL-2 production, T-cell receptor signaling pathway, and other immune regulatory pathways. In conclusion, our study shows that subcutaneous BCG administration in pre-weaned calves can induce innate immune memory in the form of trained immunity in γδ T cells. This memory is associated with increased chromatin accessibility of innate immune response-related genes, thereby inducing a functional trained immune response evidenced by increased IL-6 and TNF-α cytokine production.

XL-DNase-Seq: Footprinting Analysis of Dynamic Transcription Factors.

We have developed a novel method for genomic footprinting of transcription factors (TFs) that detects potential gene regulatory relationships from DNase-seq data at the nucleotide level. We introduce an assay termed cross-link (XL)-DNase-seq, designed to capture chromatin interactions of dynamic TFs. A mild cross-linking step in XL-DNase-seq improves the detection of DNase-based footprints of dynamic TFs. The footprint strengths and detectability depend on an optimal cross-linking procedure. This method may help extract novel gene regulatory circuits involving previously undetectable TFs. The XL-DNase-seq method is illustrated here for activated mouse macrophage-like cells, which share several features with inflammatory macrophages.

Epigenetic insights into GABAergic development in Dravet Syndrome iPSC and therapeutic implications.

Dravet syndrome (DS) is a devastating early-onset refractory epilepsy syndrome caused by variants in the SCN1A gene. A disturbed GABAergic interneuron function is implicated in the progression to DS but the underlying developmental and pathophysiological mechanisms remain elusive, in particularly at the chromatin level. Induced pluripotent stem cells (iPSCs) derived from DS cases and healthy donors were used to model disease-associated epigenetic abnormalities of GABAergic development. Chromatin accessibility was assessed at multiple time points (Day 0, Day 19, Day 35, and Day 65) of GABAergic differentiation. Additionally, the effects of the commonly used anti-seizure drug valproic acid (VPA) on chromatin accessibility were elucidated in GABAergic cells. The distinct dynamics in the chromatin profile of DS iPSC predicted accelerated early GABAergic development, evident at D19, and diverged further from the pattern in control iPSC with continued differentiation, indicating a disrupted GABAergic maturation. Exposure to VPA at D65 reshaped the chromatin landscape at a variable extent in different iPSC-lines and rescued the observed dysfunctional development of some DS iPSC-GABA. The comprehensive investigation on the chromatin landscape of GABAergic differentiation in DS-patient iPSC offers valuable insights into the epigenetic dysregulations associated with interneuronal dysfunction in DS. Moreover, the detailed analysis of the chromatin changes induced by VPA in iPSC-GABA holds the potential to improve the development of personalized and targeted anti-epileptic therapies.

Current genomic deep learning models display decreased performance in cell type specific accessible regions.

Background: A number of deep learning models have been developed to predict epigenetic features such as chromatin accessibility from DNA sequence. Model evaluations commonly report performance genome-wide; however, cis regulatory elements (CREs), which play critical roles in gene regulation, make up only a small fraction of the genome. Furthermore, cell type specific CREs contain a large proportion of complex disease heritability. Results: We evaluate genomic deep learning models in chromatin accessibility regions with varying degrees of cell type specificity. We assess two modeling directions in the field: general purpose models trained across thousands of outputs (cell types and epigenetic marks), and models tailored to specific tissues and tasks. We find that the accuracy of genomic deep learning models, including two state-of-the-art general purpose models - Enformer and Sei - varies across the genome and is reduced in cell type specific accessible regions. Using accessibility models trained on cell types from specific tissues, we find that increasing model capacity to learn cell type specific regulatory syntax - through single-task learning or high capacity multi-task models - can improve performance in cell type specific accessible regions. We also observe that improving reference sequence predictions does not consistently improve variant effect predictions, indicating that novel strategies are needed to improve performance on variants. Conclusions: Our results provide a new perspective on the performance of genomic deep learning models, showing that performance varies across the genome and is particularly reduced in cell type specific accessible regions. We also identify strategies to maximize performance in cell type specific accessible regions.

Genome-wide chromatin accessibility reveals transcriptional regulation of heterosis in inter-subspecific hybrid rice.

The utilization of rice heterosis is essential for ensuring global food security; however, its molecular mechanism remains unclear. In this study, comprehensive analyses of accessible chromatin regions (ACRs), DNA methylation, and gene expression in inter-subspecific hybrid and its parents were performed to determine the potential role of chromatin accessibility in rice heterosis. The hybrid exhibited abundant ACRs, in which the gene ACRs and proximal ACRs were directly related to transcriptional activation rather than the distal ACRs. Regarding the dynamic accessibility contribution of the parents, paternal ZHF1015 transmitted a greater number of ACRs to the hybrid. Accessible genotype-specific target genes were enriched with overrepresented transcription factors, indicating a unique regulatory network of genes in the hybrid. Compared with its parents, the differentially accessible chromatin regions with upregulated chromatin accessibility were much greater than those with downregulated chromatin accessibility, reflecting a stronger regulation in the hybrid. Furthermore, DNA methylation levels were negatively correlated with ACR intensity, and genes were strongly affected by CHH methylation in the hybrid. Chromatin accessibility positively regulated the overall expression level of each genotype. ACR-related genes with maternal Z04A-bias allele-specific expression tended to be enriched during carotenoid biosynthesis, whereas paternal ZHF1015-bias genes were more active in carbohydrate metabolism. Our findings provide a new perspective on the mechanism of heterosis based on chromatin accessibility in inter-subspecific hybrid rice.