Biology subjects

Robinson, M. D.

Publications and source records attributed to Robinson, M. D..

At least 37 records · Page 2Linked to original sources

bioRxiv2025

Bulk-based hypothesis weighing to increase power in single-cell differential expression analysis

Due to the costs of single-cell sequencing, sample sizes are often relatively limited, sometimes leading to poorly reproducible results. In many contexts, however, larger bulk RNAseq data is available for the same conditions or experimental paradigm, which can be used as additional evidence of a generalizable differential expression pattern. Here, we show how such data can be used, via bulk-based hypothesis weighing (bbhw), to increase the power and robustness of single-cell differential state analysis. We find that all methods improve performance, with the best results obtained by applying a grouped Benjamini-Hochberg procedure on bins based on proportion-adjusted significance (PAS). These methods are implemented in the muscat package, and should be applicable to a broader range of scenarios.

bioinformatics↗

bioRxiv2025

BiCLUM: Bilateral Contrastive Learning for Unpaired Single-Cell Multi-Omics Integration

The integration of single-cell multi-omics data provides a powerful approach for understanding the complex interplay between different molecular modalities, such as RNA expression, chromatin accessibility and protein abundance, measured through assays like scRNA-seq, scATAC-seq and CITE-seq, at single-cell resolution. However, most existing single-cell technologies focus on individual modalities, limiting a comprehensive understanding of their interconnections. Integrating such diverse and often unpaired datasets remains a challenging task due to unknown cell correspondences across distinct feature spaces and limited insights into cell- type-specific activities in non-scRNA-seq modalities. In this work, we propose BiCLUM, a Bilateral Contrastive Learning approach for Unpaired single-cell Multi-omics integration, which simultaneously enforces cell-level and feature-level alignment across modalities. BiCLUM first transforms one modality, such as scATAC-seq, into the data space of another modality, such as scRNA-seq, using prior genomic knowledge. It then learns cell and gene embeddings simultaneously through a bilateral contrastive learning framework, incorporating both cell-level and feature-level contrastive losses. We evaluated BiCLUM on aligning gene expression with chromatin accessibility via three paired RNA-ATAC multi-omics datasets, as well as gene expression with protein expression via three CITE-seq datasets. The results demonstrate that BiCLUM either outperforms or is at least comparable to existing integration methods, excelling in both visualization and quantitative metrics. Furthermore, BiCLUM preserves the biological relevance of the integrated data, making it a potential powerful tool for downstream biological analysis, such as cell type identification and pathway exploration.

bioinformatics↗

bioRxiv2024

On metrics for subpopulation detection in single-cell and spatial omics data

Benchmarks are crucial to understanding the strengths and weaknesses of the growing number of tools for single-cell and spatial omics analysis. A key task is to distinguish subpopulations within complex tissues, where evaluation typically relies on external clustering validation metrics. Different metrics often lead to inconsistencies between rankings, highlighting the importance of understanding the behavior and biological implications of each metric. In this work, we provide a framework for systematically understanding and selecting validation metrics for single-cell data analysis, addressing tasks such as creating cell embeddings, constructing graphs, clustering, and spatial domain detection. Our discussion centers on the desirable properties of metrics, focusing on biological relevance and potential biases. Using this framework, we not only analyze existing metrics, but also develop novel ones. Delving into domain detection in spatial omics data, we develop new external metrics tailored to spatially-aware measurements. Additionally, a Bioconductor R package, poem, implements all the metrics discussed. While we focus on single-cell omics, much of the discussion is of broader relevance to other types of high-dimensional data.

bioinformatics↗

bioRxiv2024

Disentangling cell type and state transcriptional programs

Single-cell omics approaches profile molecular constituents of individual cells. Replicated multi-condition experiments in particular aim at studying how the molecular makeup and composition of cell subpopulations changes at the sample-level. Two main approaches have been proposed for these tasks: firstly, cluster-based methods that group cells into (non-overlapping) subpopulations based on their molecular profiles and, secondly, cluster-free but neighborhood-based methods that identify (overlapping) groups of cells in consideration of cross-condition changes. In either approach, discrete cell groups are subjected to differential testing across conditions; and, a low-dimensional cell embedding, which is in turn derived from a subset of selected features, is required to delineate subpopulations or neighborhoods. We hypothesized that decoupling differences in cell type (i.e., between subpopulations) and cell state (i.e., between conditions) for feature selection would yield an embedding space that captures different aspects of cellular heterogeneity. And, that type-not-state embeddings would arrive at differential testing results that are more comparable between cluster- and neighborhood-based differential testing approaches. Our study leverages a simulation framework with competing type and state effects, as well as an experimental dataset, to evaluate a set of feature scoring and selection strategies, and to compare results from downstream differential analyses.

bioinformatics↗

bioRxiv2024

RoCK and ROI: Single-cell transcriptomics with multiplexed enrichment of selected transcripts and region-specific sequencing

Various tools have been developed to reliably identify, trace and analyze single cells in complex tissues. In recent years, these technologies have been combined with transcriptomic profiling approaches to explore molecular mechanisms that drive development, health, and disease. However, current methods still fall short of profiling single cell transcriptomes comprehensively, with one major challenge being high non-detection rates of specific transcripts and transcript regions. Such information is often crucial to understanding the biology of cells or tissues and includes lowly expressed transcripts, sequence variations and exon junctions. Here, we developed a scRNAseq workflow, RoCK and ROI (Robust Capture of Key transcripts and Regions Of Interest), that tackles these limitations. RoCKseq uses targeted capture to enrich for key transcripts, thereby supporting the detection and identification of cell types and complex phenotypes in scRNAseq experiments. ROIseq directs a subset of reads to a specific region of interest via selective priming to ensure detection. Importantly, RoCK and ROI guarantees efficient retrieval of specific sequence information without compromising overall single cell transcriptome information and our workflow is supported by a novel bioinformatics pipeline to analyze the multimodal information. RoCK and ROI represents a significant enhancement over non-targeted single cell sequencing, particularly when cell categorization depends on transcripts that are missed in standard scRNAseq experiments. In addition, it also allows exploration of biological questions that require assessment of specific sequence elements along the targets to be addressed.

genomics↗

bioRxiv2024

On the identification of differentially-active transcription factors from ATAC-seq data

ATAC-seq has emerged as a rich epigenome profiling technique, and is commonly used to identify Transcription Factors (TFs) underlying given phenomena. A number of methods can be used to identify differentially-active TFs through the accessibility of their DNA-binding motif, however little is known on the best approaches for doing so. Here we benchmark several such methods using a combination of curated datasets with various forms of short-term perturbations on known TFs, as well as semi-simulations. We include both methods specifically designed for this type of data as well as some that can be repurposed for it. We also investigate variations to these methods, and identify three particularly promising approaches (a chromVAR-limma workflow with critical adjustments, monaLisa and a combination of GC smooth quantile normalization and multivariate modeling). We further investigate the specific use of nucleosome-free fragments, the combination of top methods, and the impact of technical variation. Finally, we illustrate the use of the top methods on a novel dataset to characterize the impact on DNA accessibility of TRAnscription Factor TArgeting Chimeras (TRAFTAC), which can deplete TFs - in our case NFkB - at the protein level. Author summaryTranscription factors regulate gene expression by binding sites in the genome that often harbor a specific DNA motif. The collective accessibility of these motif-matching regions, measured by technologies such as ATAC-seq, can be used to infer the activity of the corresponding transcription factors. Here we use curated datasets of 11 TF-specific perturbations as well as 116 semi-simulated datasets to benchmark various methods for identifying factors that differ in activity between experimental conditions. We investigate important variations in the analysis and make recommendations pertaining to such analysis. Finally, we illustrate the application of the top methods to characterize the effects of a novel method for perturbing transcription factors at the protein level.

bioinformatics↗

bioRxiv2024

TTF1 control of LncRNA synthesis and cell growth delineates a tumour suppressor pathway acting directly on the ribosomal RNA Genes

The tumour suppressor p14/19ARF regulates ribosomal RNA (rRNA) synthesis by controlling the nucleolar localization of Transcription Termination Factor 1 (TTF1). However, the role played by TTF1 in regulating the rRNA genes and in potentially controlling growth has remained unclear. We now show that TTF1 expression regulates cell growth by determining the cellular complement of ribosomes. Unexpectedly, it achieves this by acting as a "roadblock" to synthesis of the non-coding LncRNA and pRNA that we show are generated from the "Spacer Promoter" duplications present upstream of the 47S pre-rRNA promoter on the mouse and human ribosomal RNA genes. Unexpectedly, the endogenous generation of these non-coding RNAs does not induce CpG methylation or gene silencing. Rather, it acts in cis to suppress 47S preinitiation complex formation and hence de novo pre-rRNA synthesis by a mechanism reminiscent of promoter interference or occlusion. Taken together, our data delineate a pathway from p19ARF to cell growth suppression via the regulation of ribosome biogenesis by non-coding RNAs and validate a key cellular growth law in mammalian cells.

molecular biology↗

bioRxiv2023

DifferentialRegulation: a Bayesian hierarchical approach to identify differentially regulated genes

MotivationAlthough transcriptomics data is typically used to analyse mature spliced mRNA, recent attention has focused on jointly investigating spliced and unspliced (or precursor-) mRNA, which can be used to study gene regulation and changes in gene expression production. Nonetheless, most methods for spliced/unspliced inference (such as RNA velocity tools) focus on individual samples, and rarely allow comparisons between groups of samples (e.g., healthy vs. diseased). Furthermore, this kind of inference is challenging, because spliced and unspliced mRNA abundance is characterized by a high degree of quantification uncertainty, due to the prevalence of multi-mapping reads, i.e., reads compatible with multiple transcripts (or genes), and/or with both their spliced and unspliced versions. ResultsHere, we present DifferentialRegulation, a Bayesian hierarchical method to discover changes between experimental conditions with respect to the relative abundance of unspliced mRNA (over the total mRNA). We model the quantification uncertainty via a latent variable approach, where reads are allocated to their gene/transcript of origin, and to the respective splice version. We designed several benchmarks where our approach shows good performance, in terms of sensitivity and error control, versus state-of-the-art competitors. Importantly, our tool is flexible, and works with both bulk and single-cell RNA-sequencing data. Availability and implementationDifferentialRegulation is distributed as a Bioconductor R package.

bioinformatics↗

bioRxiv2023

Benchmarking computational methods for single-cell chromatin data analysis

Single-cell chromatin accessibility assays, such as scATAC-seq, are increasingly employed in individual and joint multi-omic profiling of single cells. As the accumulation of scATAC-seq and multi-omics datasets continue, challenges in analyzing such sparse, noisy, and high-dimensional data become pressing. Specifically, one challenge relates to optimizing the processing of chromatin-level measurements and efficiently extracting information to discern cellular heterogeneity. This is of critical importance, since the identification of cell types is a fundamental step in current single-cell data analysis practices. We benchmarked 8 feature engineering pipelines derived from 5 recent methods to assess their ability to discover and discriminate cell types. By using 10 metrics calculated at the cell embedding, shared nearest neighbor graph, or partition levels, we evaluated the performance of each method at different data processing stages. This comprehensive approach allowed us to thoroughly understand the strengths and weaknesses of each method and the influence of parameter selection. Our analysis provides guidelines for choosing analysis methods for different datasets. Overall, feature aggregation, SnapATAC, and SnapATAC2 outperform latent semantic indexing-based methods. For datasets with complex cell-type structures, SnapATAC and SnapATAC2 are preferred. With large datasets, SnapATAC2 and ArchR are most scalable.

bioinformatics↗

bioRxiv2023

Validation of hypermethylated DNA regions found in colorectal cancers as potential aging-independent biomarkers of precancerous colorectal lesions

BackgroundWe previously identified 16,772 colorectal cancer-associated hypermethylated DNA regions that were also detectable in precancerous colorectal lesions (preCRCs) and unrelated to normal mucosal aging. We have now conducted a study to validate 990 of these differentially methylated DNA regions (DMR) in a new series of preCRCs. MethodsWe used targeted bisulfite sequencing to validate these 990 potential biomarkers in 59 preCRC tissue samples (41 conventional adenomas, 18 sessile serrated lesions), each with a patient-matched normal mucosal sample. Based on differential DNA methylation tests, a panel of (candidate) DMRs was chosen on a subset of the (our) cohort and validated on the remaining part of our cohort and (two) further publicly available datasets with respect to their stratifying potential between preCRCs and normal mucosa. ResultsStrong statistical significance for the difference in methylation levels was observed across the full set of 990 investigated DMRs. From these, a selected candidate panel of 30 DMRs correctly identified 58/59 tumors (area under the receiver operating curve: 0.998). ConclusionsThese validated DNA hypermethylation markers can be exploited to develop more accurate noninvasive colorectal tumor screening assays.

genomics↗

bioRxiv2023

DESpace: spatially variable gene detection via differential expression testing of spatial clusters

MotivationSpatially resolved transcriptomics (SRT) enables scientists to investigate spatial context of mRNA abundance, including identifying spatially variable genes (SVGs), i.e., genes whose expression varies across the tissue. Although several methods have been proposed for this task, native SVG tools cannot jointly model biological replicates, or identify the key areas of the tissue affected by spatial variability. ResultsHere, we introduce DESpace, a framework, based on an original application of existing methods, to discover SVGs. In particular, our approach inputs all types of SRT data, summarizes spatial information via spatial clusters, and identifies spatially variable genes by performing differential gene expression testing between clusters. Furthermore, our framework can identify (and test) the main cluster of the tissue affected by spatial variability; this allows scientists to investigate spatial expression changes in specific areas of interest. Additionally, DESpace enables joint modelling of multiple samples (i.e., biological replicates); compared to inference based on individual samples, this approach increases statistical power, and targets SVGs with consistent spatial patterns across replicates. Overall, in our benchmarks, DESpace displays good true positive rates, controls for false positive and false discovery rates, and is computationally efficient. Availability and implementationDESpace is freely distributed as a Bioconductor R package.

bioinformatics↗

bioRxiv2023

TAR Syndrome-associated Rbm8a deficiency causes hematopoietic defects and attenuates Wnt/PCP signaling

Defects in blood development frequently occur among syndromic congenital anomalies. Thrombocytopenia-Absent Radius (TAR) syndrome is a rare congenital condition with reduced platelets (hypomegakaryocytic thrombocytopenia) and forelimb anomalies, concurrent with more variable heart and kidney defects. TAR syndrome associates with hypomorphic gene function for RBM8A/Y14 that encodes a component of the exon junction complex involved in mRNA splicing, transport, and nonsense-mediated decay. How perturbing a general mRNA-processing factor causes the selective TAR Syndrome phenotypes remains unknown. Here, we connect zebrafish rbm8a perturbation to early hematopoietic defects via attenuated non-canonical Wnt/Planar Cell Polarity (PCP) signaling that controls developmental cell re-arrangements. In hypomorphic rbm8a zebrafish, we observe a significant reduction of cd41-positive thrombocytes. rbm8a-mutant zebrafish embryos accumulate mRNAs with individual retained introns, a hallmark of defective nonsense-mediated decay; affected mRNAs include transcripts for non-canonical Wnt/PCP pathway components. We establish that rbm8a-mutant embryos show convergent extension defects and that reduced rbm8a function interacts with perturbations in non-canonical Wnt/PCP pathway genes wnt5b, wnt11f2, fzd7a, and vangl2. Using live-imaging, we found reduced rbm8a function impairs the architecture of the lateral plate mesoderm (LPM) that forms hematopoietic, cardiovascular, kidney, and forelimb skeleton progenitors as affected in TAR Syndrome. Both mutants for rbm8a and for the PCP gene vangl2 feature impaired expression of early hematopoietic/endothelial genes including runx1 and the megakaryocyte regulator gfi1aa. Together, our data propose aberrant LPM patterning and hematopoietic defects as consequence of attenuated non-canonical Wnt/PCP signaling upon reduced rbm8a function. These results also link TAR Syndrome to a potential LPM origin and a developmental mechanism. HIGHLIGHTSO_LIZebrafish mutants for the TAR Syndrome gene rbm8a show thrombocyte reduction C_LIO_LIAttenuated Rbm8a function results in retained introns in mRNAs encoding PCP components C_LIO_LIEarly PCP defects result in lateral plate mesoderm anomalies and hematopoietic defects C_LIO_LIPCP anomalies selectively impact cell fate patterning C_LI

developmental biology↗

bioRxiv2022

Meta-analysis of (single-cell method) benchmarks reveals the need for extensibility and interoperability

Computational methods represent the lifeblood of modern molecular biology. Benchmarking is important for all methods, but with a focus here on computational methods, benchmarking is critical to dissect important steps of analysis pipelines, formally assess performance across common situations as well as edge cases, and ultimately guide users on what tools to use. Benchmarking can also be important for community building and advancing methods in a principled way. We conducted a meta-analysis of recent single-cell benchmarks to summarize the scope, extensibility, neutrality, as well as technical features and whether best practices in open data and reproducible research were followed. The results highlight that while benchmarks often make code available and are in principle reproducible, they remain difficult to extend, for example, as new methods and new ways to assess methods emerge. In addition, embracing containerization and workflow systems would enhance reusability of intermediate benchmarking results, thus also driving wider adoption.

bioinformatics↗

bioRxiv2022

A comprehensive single-cell atlas of freshly dissociated human synovium in inflammatory arthritis with an optimized dissociation protocol for prospective fresh synovial biopsy collection

Single-cell RNA-sequencing is advancing our understanding of synovial pathobiology in inflammatory arthritis. Here, we optimized the protocol for the dissociation of fresh synovial biopsies and created a reference single-cell map of fresh human synovium in inflammatory arthritis. We utilized the published method for dissociating cryopreserved synovium and optimized it for dissociating small fresh synovial biopsies. The optimized protocol enabled the isolation of a good yield of consistently highly viable cells, minimizing the dropout rate of prospectively collected biopsies. Our reference synovium map comprised over 100000 unsorted single-cell profiles from 25 synovial tissues of patients with inflammatory arthritis. Synovial cells formed 11 lymphoid, 15 myeloid and 16 stromal cell clusters, including IFITM2+ synovial neutrophils. Using this reference map, we successfully annotated published synovial scRNA-seq datasets. Our dataset uncovered endothelial cell diversity and identified SOD2highSAA1+SAA2+ and SERPINE1+COL5A3+ fibroblast clusters, expressing genes linked to cartilage breakdown (SDC4) and extracellular matrix remodelling (LOXL2, TGFBI, TGFB1), respectively. We broadened the characterization of tissue resident FOLR2+COLEC12high and LYVE1+SLC40A1+ macrophages, inferring their extracellular matrix sensing and iron recycling activities. Our research brings an efficient synovium dissociation protocol and a reference annotation resource of fresh human synovium, while expanding the knowledge about synovial cell diversity in inflammatory arthritis.

genomics↗

bioRxiv2022

The structure of yeast Npl3 bound to RNA reveals a cooperative sequence-specific recognition and an RNA chaperone role in splicing

The conserved SR-like protein Npl3 promotes splicing of diverse pre-mRNAs. However, the RNA sequence(s) recognized by the RNA Recognition Motifs (RRM1 & RRM2) of Npl3 during the splicing reaction remain elusive. Here, we developed a split-iCRAC approach in yeast to uncover the consensus sequence bound to each RRM. High-resolution NMR structures show that RRM2 recognizes a 5-GNGG-3 motif leading to an unusual mille-feuille topology. These structures also reveal how RRM1 preferentially interacts with a CC-dinucleotide upstream of this motif, and how the inter-RRM linker and the region C-terminal to RRM2 contributes to cooperative RNA-binding. Structure-guided functional studies show that Npl3 genetically interacts with U2 snRNP specific factors and we provide evidence that Npl3 melts U2 snRNA stem-loop I, a prerequisite for U2/U6 duplex formation within the catalytic center of the Bact spliceosomal complex. Thus, our findings provide insights into an unanticipated RNA chaperoning role for Npl3 during spliceosome active site formation.

biophysics↗

bioRxiv2022

Supervised spatial inference of dissociated single-cell data with SageNet

Spatially-resolved transcriptomics uncovers patterns of gene expression at supercellular, cellular, or subcellular resolution, providing insights into spatially variable cellular functions, diffusible morphogens, and cell-cell interactions. However, for practical reasons, multiplexed single cell RNA-sequencing remains the most widely used technology for profiling transcriptomes of single cells, especially in the context of large-scale anatomical atlassing. Devising techniques to accurately predict the latent physical positions as well as the latent cell-cell proximities of such dissociated cells, represents an exciting and new challenge. Most of the current approaches rely on an autocorrelation assumption, i.e., cells with similar transcriptomic profiles are located close to each other in physical space and vice versa. However, this is not always the case in native biological contexts due to complex morphological and functional patterning. To address this challenge, we developed SageNet, a graph neural network approach that spatially reconstructs dissociated single cell data using one or more spatial references. SageNet first estimates a gene-gene interaction network from a reference spatial dataset. This informs the structure of the graph on which the graph neural network is trained to predict the region of dissociated cells. Finally, SageNet produces a low-dimensional embedding of the query dataset, corresponding to the reconstructed spatial coordinates of the dissociated tissue. Furthermore, SageNet reveals spatially informative genes by extracting the most important features from the neural network model. We demonstrate the utility and robust performance of SageNet using molecule-resolved seqFISH and spot-based Spatial Transcriptomics reference datasets as well as dissociated single-cell data, across multiple biological contexts. SageNet is provided as an open-source python software package at https://github.com/MarioniLab/SageNet.

bioinformatics↗

bioRxiv2022

Single nuclei RNAseq stratifies multiple sclerosis patients into three distinct white matter glia responses

The lack of understanding of the cellular and molecular basis of clinical and genetic heterogeneity in progressive multiple sclerosis (MS) has hindered the search for new effective therapies. Here, to address this gap, we analysed 632,000 single nuclei RNAseq profiles of 156 brain tissue samples, comprising white matter (WM) lesions, normal appearing WM, grey matter (GM) lesions and normal appearing GM from 54 MS patients and 26 controls. We observed the expected changes in overall neuronal and glial numbers previously described within the classical lesion subtypes. We found highly cell type-specific gene expression changes in MS tissue, with distinct differences between GM and WM areas, confirming different pathologies. However, surprisingly, we did not observe distinct gene expression signatures for the classical different WM lesion types, rather a continuum of change. This indicates that classical lesion characterization better reflects changes in cell abundance than changes in cell type gene expression, and indicates a global disease effect. Furthermore, the major biological determinants of variability in gene expression in MS WM samples relate to individual patient effects, rather than to lesion types or other metadata. We identify four subgroups of MS patients with distinct WM glial gene expression signatures and patterns of oligodendrocyte stress and/or maturation, suggestive of engagement of different pathological processes, with an additional more variable regenerative astrocyte signature. The discovery of these patterns, which were also found in an independent MS patient cohort, provides a framework to use molecular biomarkers to stratify patients for optimal therapeutic approaches for progressive MS, significantly advances our mechanistic understanding of progressive MS, and highlights the need for precision-medicine approaches to address heterogeneity among MS patients.

neuroscience↗

bioRxiv2022

Gapless provides combined scaffolding, gap filling and assembly correction with long reads

Continuity, correctness and completeness of genome assemblies are important for many biological projects. Long reads represent a major driver towards delivering high-quality genomes, but not everybody can achieve the necessary coverage for good long-read-only assemblies. Therefore, improving existing assemblies with low-coverage long reads is a promising alternative. The improvements include correction, scaffolding and gap filling. However, most tools perform only one of these tasks and the useful information of reads that supported the scaffolding is lost when running separate programs successively. Therefore, we propose a new tool for combined execution of all three tasks using PacBio or Oxford Nanopore reads. gapless is available at: https://github.com/schmeing/gapless.

bioinformatics↗