Genome Biology最新文献

RNA-seq data analysis relies on many different tools, each tailored to specific applications and coming with unique assumptions and limitations. Indeed, tools for differential transcript usage or rare disease diagnosis through splicing and expression outliers, either lack performance, discard information, or do not scale to large datasets. We show that replacing normalization offsets unlocks bulk RNA-seq tools for differential usage and aberrant splicing, providing a single framework for various short- and long-read applications. We then introduce saseR, a tool for prioritizing expression and usage outliers that is much faster than state-of-the-art methods, and significantly outperforms these for aberrant splicing detection.

Introduction: Recently, de novo transcriptome assembly methods have been developed to utilise long-read data in cases where a reference genome is unavailable, such as in non-model organisms. Despite the potential of these tools, there remains a lack of benchmarking and established protocols for optimal reference-free, long-read transcriptome assembly and differential expression analysis.

Results: Here, we evaluate the long-read de novo transcriptome assembly tools, RATTLE, RNA-Bloom2 and isONform, and compare their performance to one of the leading short-read assemblers, Trinity. We assess various metrics across a range of datasets, which include simulated data and spike-in sequin transcripts, where ground truth is known, and real data from human and pea (Pisum sativum) samples, using a reference-based approach to define truth. To represent contemporary analysis scenarios, the datasets cover depths from 6 to 60 million reads, Oxford Nanopore Technologies (ONT) cDNA, ONT direct RNA and Pacific Biosciences (PacBio) 10 × single-cell sequencing. Critically, we assess the downstream impact of assembly choice on the detection of differential gene and transcript expression.

Conclusions: Our results confirm that long reads generate longer assembled transcripts than short-reads for reference-free analysis, though limitations remain compared to reference-guided approaches, and suggest scope for improved accuracy and reduced redundancy. Of the de novo pipelines, RNA-Bloom2, coupled with Corset for transcript clustering, was the best performing in terms of both accuracy and computational efficiency. Our findings offer guidance when selecting the most effective strategy for long-read differential expression analysis, when a high-quality reference genome is unavailable.

Background: Brassica napus (B. napus) is globally important oilseed crop, yet traditional approaches for phenotyping of seed traits are labor-intensive and destructive.

Results: Here, we establish a non-destructive analytical framework integrating hyperspectral imaging (HSI) with machine learning for characterizing seed-related traits. We collect HSI data from seeds of 393 B. napus accessions over two consecutive years, generating 1,944 spectral indices per sample. We identify significant correlations between 1,293 hyperspectral indices and 956 seed metabolites. Flavonoid metabolites exhibit the most consistent interannual correlations with hyperspectral indices. Systematic benchmarking of 19 machine learning algorithms identifies nine optimal models for metabolite prediction, with 73.44% of metabolites achieving significant associations. Hyperspectral indices effectively predict nine key seed-related traits, including oil content, seed coat content, glucosinolate content and six fatty acid components. Genome-wide association studies (GWAS) of hyperspectral indices uncover three stable quantitative trait loci (QTL) hotspots, qHSI.hotA09, qHSI.hotA05 and qHSI.hotC05, that co-localize with QTLs for seed oil and seed coat content. Integration of GWAS with POCKET prioritization identifies BnaA09.MYB52 and BnaC05.PMT6 as candidate genes for the hotspots, qHSI.hotA09 and qHSI.hotC05, respectively. Functional validation using mutants demonstrates that both genes significantly influence seed flavonoid metabolites and hyperspectral profiles. BnaPMT6 is characterized as a novel positive regulator of seed coat content, similar to BnaMYB52.

Conclusions: This study establishes a novel, non-destructive approach for seed traits and metabolite assessment in B. napus seeds. It also provides a theoretical foundation and genetic basis for breeding of B. napus varieties with high oil content and improved nutritional quality.

Detecting senescent cells from single-cell RNA-seq data remains challenging due to the weak and non-specific expression of canonical markers. Here, we demonstrate that simple expansion of these low-signal marker sets does not improve detection accuracy. To address this limitation, we develop ICE (Imputation-based Cell Enrichment), a computational framework that integrates expression imputation with marker refinement. ICE improves the detection of senescent cells in pancreatic β cells and microglia from Alzheimer's disease samples. This tool enables reliable identification of senescence-associated cell populations, facilitating more detailed analyses of their heterogeneity and temporal dynamics across human tissues and disease contexts.

Elucidating transcription factor (TF) function is essential for advancing our understanding and manipulation of the mechanisms that orchestrate gene expression programs underlying plant growth, development, and resilience. Defining the genomic binding sites of a TF-its cistrome-is particularly critical, as it reveals where TF activity can occur and, when integrated with gene expression and chromatin landscape data, delineates the full scope of TF function. In this review, we highlight the biological factors that shape plant cistromes and, importantly, the potential alterations in cistrome composition that may arise during experimental mapping. We further emphasize recent methodological advances now available to the plant science community and outline future directions for the emerging field of plant cistromics.

Antifungal peptides (AFPs) are crucial for plant defense against biotic stress. Yet, no artificial intelligence tool specifically classifies plant AFPs. To fill this gap, we develop FungiGuard, which integrates Random Forest, Long Short-Term Memory, and attention mechanisms to identify AFPs using functionally annotated plant small peptides. FungiGuard outperforms existing generalized AFP model in classifying plant AFPs, and detects candidate AFPs in Arabidopsis, wheat, rice, and maize. It also discovers novel AFPs through randomly generated sequences. Experimental validation confirms the antifungal activity of candidate AFP against Botrytis cinerea. This tool deepens plant AFP understanding and facilitates novel AFP discovery.

Background: Glucocorticoids are corticosteroid hormones that are commonly used for treating systemic inflammatory diseases and acute infections. Immunosuppressive effects of glucocorticoids have been studied in many cell types, particularly macrophages and T cells. Despite the importance and abundance of neutrophils in the human immune system, glucocorticoid responses remain understudied in neutrophils.

Results: Here, we perform quantitative mass spectrometry-based proteomics of primary neutrophils and neutrophil-like cells differentiated from human HL-60 promyelocyte cells. Primary neutrophils exhibited CK2 kinase activation and increase phosphorylation of HSP90 following 2-h incubation, highlighting potential effects of short-term ex vivo handling. Proteome and flow cytometry analysis show that neutrophil-like cells share features of neutrophils. Quantitative proteomics and phosphoproteomics of neutrophil-like cells treated with two synthetic glucocorticoid compounds, the clinical drugs dexamethasone and prednisolone, identify higher numbers of significantly regulated proteins and phosphosites compared to parental HL-60 cells. Glucocorticoid treatments modulated toll-like receptor signaling and CXCR4 serine phosphorylation. In addition, we identify RIPOR2 as a glucocorticoid-regulated protein associated with Rho GTPase signaling networks and actin cytoskeletal remodeling in neutrophils and neutrophil-like cells, though its exact functional role requires further investigation.

Conclusions: Our results not only reveal unconventional regulatory mechanisms of glucocorticoids in the human immune system but also provide valuable resources for discovering novel glucocorticoid-responsive protein targets in neutrophils.

Background: Aging is a multi-modal process, leaving distinct molecular signatures across the epigenome. DNA methylation is among the most robust biomarkers of biological aging, yet most studies assume linear age relationships and analyze mixed-sex cohorts, overlooking known sex differences. Such approaches risk obscuring critical nonlinear transitions and sex-specific trajectories.

Results: We develop SNITCH, a computational framework to detect complex nonlinear methylation trajectories and disentangle shared from sex-divergent patterns. Applied to the array-derived whole-blood methylomes from 252 females and 246 males (ages 19-90 years), SNITCH reveals convergent and divergent epigenetic aging pathways independent of immune cell composition. Nonlinear trajectories are enriched for developmental transcription factor motifs, including NF1/CTF and REST, with known oncogenic roles. Importantly, a female-specific nonlinear cluster is prospectively associated with cancer onset and systemic inflammation in an independent cohort, nominating clinically relevant biomarkers. We replicate the analysis in an additional cohort and highlight consistent nonlinear trajectories.

Conclusions: Our results uncover sex-specific, nonlinear aging programs that capture the dynamics of epigenetic change beyond linear models. These findings provide potential candidate biomarkers for early disease risk and advance understanding of how aging trajectories diverge between sexes.

Dimensionality reduction is routinely applied to single-cell transcriptomic data to improve interpretability, remove noise and redundancy, and enable visualization. Most existing methods aim at preserving the most prominent data properties, which can lead to omission of rare but important signals. Here we propose a novel framework, SAKURA, that uses knowledge-derived genes of interest to guide dimensionality reduction, which can help cluster rare cells and separate highly similar cell subpopulations. We demonstrate the utility of our framework in identifying endocrine cell subtypes in the pancreatic islet, highly similar hematopoietic subpopulations, and rare senescent cells.

Drug combinations can improve cancer therapy by boosting efficacy, limiting dose-related toxicity, and delaying resistance. We present UniSyn, an interpretable multi-modal deep learning framework that transfers knowledge from monotherapy responses to enhance drug-synergy prediction. Through hybrid attention-based integration of drug and cell-line features, UniSyn supports multi-task learning and yields mechanistic insights. It generalizes robustly to unseen drug pairs and cell types, maintaining consistent performance across multiple synergy scoring metrics. Applied at scale to tumor cell lines, UniSyn captures context-specific synergy signals and prioritizes therapeutic combinations with translational potential.