Biology subjects

Svensson, V.

Publications and source records attributed to Svensson, V..

15 recordsLinked to original sources

bioRxiv2026

A single-cell RNA-seq catalog of ground truth gene coregulation

Single-cell RNA-sequencing measurements are uniquely well-poised to identify coregulated gene transcription. Coregulation should be apparent as correlation of expression levels, but at which unit to quantify expression for calculation of correlations is not clear. To enable evaluation of normalization methods for identification of coregulation we have curated and preprocessed 35 scRNA-seq samples with known regulature relationships into a catalog. These datasets contain promoter-reporter data representing known positive coregulation and B cells where isotypic mutual exclusion of {kappa} and {lambda} light chains represent known negative coregulation. As an example, we use the catalog to evaluate seven normalization methods. Different normalization methods perform better for positive or negative coregulation, and results indicate implementation choices for advanced normalization methods The catalog may serve as a resource for evaluating novel proposed normalization methods and is available at https://huggingface.co/datasets/valsv/scrna-coregulation-benchmark

bioinformatics↗

bioRxiv2026

Back to basics: Observed statistics are sufficient to predict drug responses

Predicting how cells, tissues, and patients will respond to a drug, cytokine, or genetic perturbation is central to biological and clinical reasoning. The practical goal is to estimate the analysis-ready readouts that support this reasoning: which cellular responses are context-dependent, which perturbations reveal shared or divergent mechanisms, and which observations should become the basis for the next experiment. Here we introduce Rhaister, a perturbation-response predictor that operates directly on screen-level summary statistics. By measuring just a few perturbations in a new biological context, Rhaister predicts the unmeasured perturbations by learning how response patterns vary across reference contexts. This formulation applies to both fine-grained molecular readouts, such as transcriptional responses from Tahoe-100M or other large perturbation screen, or on phenotypic endpoints. To train and apply Rhaister on pheno-typic endpoints we created Emerald Bay, a purpose-built Tahoe dataset that unifies multi-day cancer drug perturbation, pooled Mosaic tumor contexts, and paired transcriptomic response measurements*. Across these settings, Rhaister matches or exceeds substantially more expensive virtual-cell models, often achieving the highest values possible in evaluation metrics, while training in seconds and running predictions in milliseconds. On Emerald Bay, Rhaister predicts context-specific drug phenotypes from sensitivity measurements alone and improves further when including transcriptomic information. We also introduce Rhaister-O, predicting drug responses in new contexts from baseline expression alone and, to our knowledge, provides the first zero-shot model for this task. Rhaister establishes summary-statistic perturbation modeling as a fast, interpretable framework for predicting biological response across new contexts.

genomics↗

bioRxiv2026

Improving SCVI for low-count cells through self-supervised augmentation

When analyzing single-cell RNA sequencing data with SCVI, low-UMI cells typically need to be filtered, because their learned representations lack meaningful biological signal. We show that this is caused by a specific mechanism: as UMI depth decreases, the SCVI encoder maps cells towards a learned bias point, collapsing their representations regardless of cell identity. This phenomenon is distinct from classical posterior collapse driven by KL regularization. By augmenting training with binomial thinning and adding a cross-correlation loss between original and thinned cell embeddings, the encoder learns representations that preserve cell type identity, experimental condition differences, and sample-level variation at lower UMI depths, without sacrificing reconstruction quality. These modifications extend the range of usable cells, enabling analysis of cells that would typically be discarded due to low molecule counts.

bioinformatics↗

bioRxiv2025

Tahoe-x1: Scaling Perturbation-Trained Single-CellFoundation Models to 3 Billion Parameters

Foundation models have transformed natural language processing and computer vision, yet their potential in single-cell biology--particularly for complex diseases such as cancer-- remains underexplored. We present Tahoe-x1 (Tx1), a family of perturbation-trained single-cell foundation models with up to 3 billion parameters. Tx1 is pretrained on large-scale single-cell transcriptomic datasets, including the Tahoe-100M perturbation compendium, and fine-tuned for cancer-relevant tasks. Through architectural optimizations, data loader refinements, and efficient training strategies, Tx1 achieves 3-30x higher compute efficiency than prior implementations of cell-state models. Tx1 jointly learns representations of genes, cells, and compounds using a masked-expression generative objective that incorporates a drug token, enabling flexible adaptation to diverse downstream applications. We evaluate Tx1 across four key disease-relevant benchmarks: (1) prediction of overall and context-specific gene essentiality, (2) identification of genes contributing to the hallmarks of cancer, (3) cell-type classification, and (4) prediction of perturbation responses in held-out cellular contexts. Tx1 achieves state-of-the-art performance across all tasks. We release pretrained checkpoints, training code, and evaluation workflows to accelerate the development of perturbation-trained single-cell foundation models for applications in precision oncology and beyond. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=52 SRC="FIGDIR/small/683759v1_ufig1.gif" ALT="Figure 1"> View larger version (13K): org.highwire.dtl.DTLVardef@aa33d5org.highwire.dtl.DTLVardef@310a38org.highwire.dtl.DTLVardef@195f204org.highwire.dtl.DTLVardef@143f7e3_HPS_FORMAT_FIGEXP M_FIG C_FIG

systems biology↗

bioRxiv2025

Tahoe-100M: A Giga-Scale Single-Cell Perturbation Atlas for Context-Dependent Gene Function and Cellular Modeling

Building predictive models of the cell requires systematically mapping how perturbations reshape each cells state, function, and behavior. Here, we present Tahoe-100M, a giga-scale single-cell atlas of 100 million transcriptomic profiles measuring how each of 1,100 small-molecule perturbations impact cells across 50 cancer cell lines. Our high-throughput Mosaic platform, composed of a highly diverse and optimally balanced "cell village", reduces batch effects and enables parallel profiling of thousands of conditions at single-cell resolution at an unprecedented scale. As the largest single-cell dataset to date, Tahoe-100M enables artificial-intelligence (AI)-driven models to learn context-dependent functions, capturing fundamental principles of gene regulation and network dynamics. Although we leverage cancer models and pharmacological compounds to create this resource, Tahoe-100M is fundamentally designed as a broadly applicable perturbation atlas and supports deeper insights into cell biology across multiple tissues and contexts. By publicly releasing this atlas, we aim to accelerate the creation and development of robust AI frameworks for systems biology, ultimately improving our ability to predict and manipulate cellular behaviors across a wide range of applications.

genomics↗

bioRxiv2021

scvi-tools: a library for deep probabilistic analysis of single-cell omics data

AO_SCPLOWBSTRACTC_SCPLOWProbabilistic models have provided the underpinnings for state-of-the-art performance in many single-cell omics data analysis tasks, including dimensionality reduction, clustering, differential expression, annotation, removal of unwanted variation, and integration across modalities. Many of the models being deployed are amenable to scalable stochastic inference techniques, and accordingly they are able to process single-cell datasets of realistic and growing sizes. However, the community-wide adoption of probabilistic approaches is hindered by a fractured software ecosystem resulting in an array of packages with distinct, and often complex interfaces. To address this issue, we developed scvi-tools (https://scvi-tools.org), a Python package that implements a variety of leading probabilistic methods. These methods, which cover many fundamental analysis tasks, are accessible through a standardized, easy-to-use interface with direct links to Scanpy, Seurat, and Bioconductor workflows. By standardizing the implementations, we were able to develop and reuse novel functionalities across different models, such as support for complex study designs through nonlinear removal of unwanted variation due to multiple covariates and reference-query integration via scArches. The extensible software building blocks that underlie scvi-tools also enable a developer environment in which new probabilistic models for single cell omics can be efficiently developed, benchmarked, and deployed. We demonstrate this through a code-efficient reimplementation of Stereoscope for deconvolution of spatial transcriptomics profiles. By catering to both the end user and developer audiences, we expect scvi-tools to become an essential software dependency and serve to formulate a community standard for probabilistic modeling of single cell omics.

bioinformatics↗

bioRxiv2020

Single-cell transcriptomics of allo-reactive CD4+ T cells over time reveals divergent fates during gut GVHD

Acute gastrointestinal Graft-versus-Host-Disease (GVHD) is a primary determinant of mortality after allogeneic hematopoietic stem-cell transplantation (alloSCT). It is mediated by alloreactive donor CD4+ T cells that differentiate into pathogenic subsets expressing IFN{gamma}, IL-17A or GM-CSF, and is regulated by subsets expressing IL-10 and/or Foxp3. Developmental relationships between T-helper states during priming in mesenteric lymph nodes (mLN) and effector function in the GI tract remain undefined at genome-scale. We used scRNA-seq and computational modelling to create an atlas of putative differentiation pathways during GVHD. Computational trajectory inference suggested emergence of pathogenic and regulatory states along a single developmental trajectory in mLN. Importantly, we identified an unexpected second trajectory, categorised by little proliferation or cytokine expression, reduced glycolysis, and high TCF1 expression. TCF1hi cells upregulated 4{beta}7 prior to gut migration and failed to express cytokines therein. Nevertheless, they demonstrated recall potential and plasticity following secondary transplantation, including cytokine or Foxp3 expression, but reduced TCF1. Thus, scRNA-seq revealed divergence of allo-reactive CD4+ T cells into quiescent and effector states during gut GVHD, reflecting putative heterogenous priming in vivo. These findings, the first at a single-cell level during GVHD over time, can now be used to interrogate T cell differentiation in patients undergoing alloSCT.

immunology↗

bioRxiv2019

Quantifying the tradeoff between sequencing depth and cell number in single-cell RNA-seq

The allocation of a sequencing budget when designing single cell RNA-seq experiments requires consideration of the tradeoff between number of cells sequenced and the read depth per cell. One approach to the problem is to perform a power analysis for a univariate objective such as differential expression. However, many of the goals of single-cell analysis requires consideration of the multivariate structure of gene expression, such as clustering. We introduce an approach to quantifying the impact of sequencing depth and cell number on the estimation of a multivariate generative model for gene expression that is based on error analysis in the framework of a variational autoencoder. We find that at shallow depths, the marginal benefit of deeper sequencing per cell significantly outweighs the benefit of increased cell numbers. Above about 15,000 reads per cell the benefit of increased sequencing depth is minor. Code for the workflow reproducing the results of the paper is available at https://github.com/pachterlab/SBP_2019/.

genomics↗

bioRxiv2019

A curated database reveals trends in single cell transcriptomics

The more than 500 single-cell transcriptomics studies that have been published to date constitute a valuable and vast resource for biological discovery. While various \"atlas\" projects have collated some of the associated datasets, most questions related to specific tissue types, species, or other attributes of studies require identifying papers through manual and challenging literature search. To facilitate discovery with published single-cell transcriptomics data, we have assembled a near exhaustive, manually curated database of single-cell transcriptomics studies with key information: descriptions of the type of data and technologies used, along with descriptors of the biological systems studied. Additionally, the database contains summarized information about analysis in the papers, allowing for analysis of trends in the field. As an example, we show that the number of cell types identified in scRNA-seq studies is proportional to the number of cells analysed. The database is available at www.nxn.se/single-cell-studies/gui.

genomics↗

bioRxiv2019

Interpretable factor models of single-cell RNA-seq via variational autoencoders

Single cell RNA-seq makes possible the investigation of variability in gene expression among cells, and dependence of variation on cell type. Statistical inference methods for such analyses must be scalable, and ideally interpretable. We present an approach based on a modification of a recently published highly scalable variational autoencoder framework that provides interpretability without sacrificing much accuracy. We demonstrate that our approach enables identification of gene programs in massive datasets. Our strategy, namely the learning of factor models with the auto-encoding variational Bayes framework, is not domain specific and may be of interest for other applications.

bioinformatics↗

bioRxiv2019

Transcriptome Dynamics Reveals Progressive Transition from Effector to Memory in CD4+ T cells.

CD4+ T cells are repositories of immune memory, conferring enhanced immunity to many infectious agents. Studies of acute viral and bacterial infection suggest that memory CD4+ T cells develop directly from effectors. However, delineating these dynamic developmental pathways has been challenging. Here, we used high-resolution single-cell RNA-seq and temporal mixture modelling to examine the fate of Th1 and Tfh effector cells during non-lethal Plasmodium infection in mice. We observed linear Th1 and Tfh pathways towards memory, characterized by progressive halving in the numbers of genes expressed, and partial transcriptomic coalescence. Low-level persisting infection diverted but did not block these pathways. We observed in the Th1-pathway a linear transition from Th1 through a Tr1 state to TEM cells, which were then poised for Th1 re-call. The Tfh-pathway exhibited a modest Th1-signature throughout, with little evidence of Tr1 development, and co-expression of TCM and memory Tfh markers. Thus, we present a high-resolution atlas of transcriptome dynamics for naive to memory transitions in CD4+ T cells. We also defined a subset of memory-associated genes, including transcription factors Id2 and Maf, whose expression increased progressively against the background of transcriptomic quiescence. Single-cell ATAC-seq revealed substantial heterogeneity in chromatin accessibility in single effectors, which was extensively, though incompletely reset and homogenized in memory. Our data reveal that linear transitions from effector to memory occur in a progressive manner over several weeks, suggesting opportunities for manipulating CD4+ T cell memory after primary infection.\n\nHighlightsO_LIscRNA-seq reveals progressive transition from effector to memory in CD4+ T cells.\nC_LIO_LITranscriptome dynamics suggest linear not branching models for memory development.\nC_LIO_LIA subset of genes associates with gradual onset of CD4+ T cell memory.\nC_LIO_LITh1/Tfh predisposition varies among clonotypes with identical antigen-specificity.\nC_LIO_LIscATAC-seq uncovers non-coding \"memory\" elements in the genome.\nC_LI

immunology↗

bioRxiv2019

RNA velocity and protein acceleration from single-cell multiomics experiments

The simultaneous quantification of protein and RNA makes possible the inference of past, present and future cell states from single experimental snapshots. To enable such temporal analysis from multimodal single-cell experiments, we introduce an extension of the RNA velocity method that leverages estimates of unprocessed transcript and protein abundances to extrapolate cell states. We apply the model to four datasets and demonstrate consistency among landscapes and phase portraits.

bioinformatics↗

bioRxiv2019

Droplet scRNA-seq is not zero-inflated

Potential users of single cell RNA-sequencing often encounter a choice between high-throughput droplet based methods and high sensitivity plate based methods. In particular there is a widespread belief that single-cell RNA-sequencing will often fail to generate measurements for particular gene, cell pairs due to molecular inefficiencies, causing data to have an overabundance of zero-values. Investigation of published data of technical controls in droplet based single cell RNA-seq experiments demonstrates the number of zeros in the data is consistent with count statistics, indicating that over-abundances of zero-values in biological data are likely due to biological variation as opposed to technical shortcomings.

bioinformatics↗

bioRxiv2018

Single-cell transcriptomics identifies CD44 as a new marker and regulator of haematopoietic stem cells development

The endothelial to haematopoietic transition (EHT) is the process whereby haemogenic endothelium differentiates into haematopoietic stem and progenitor cells (HSPCs). The intermediary steps of this process are unclear, in particular the identity of endothelial cells that give rise to HSPCs is unknown. Using single-cell transcriptome analysis and antibody screening we identified CD44 as a new marker of EHT enabling us to isolate robustly the different stages of EHT in the aorta gonad mesonephros (AGM) region. This allowed us to provide a very detailed phenotypical and transcriptional profile for haemogenic endothelial cells, characterising them with high expression of genes related to Notch signalling, TGFbeta/BMP antagonists (Smad6, Smad7 and Bmper) and a downregulation of genes related to glycolysis and the TCA cycle. Moreover, we demonstrated that by inhibiting the interaction between CD44 and its ligand hyaluronan we could block EHT, identifying a new regulator of HSPC development.

developmental biology↗

bioRxiv2017

SpatialDE - Identification Of Spatially Variable Genes

Technological advances have enabled low-input RNA-sequencing, paving the way for assaying transcriptome variation in spatial contexts, including in tissues. While the generation of spatially resolved transcriptome maps is increasingly feasible, computational methods for analysing the resulting data are not established. Existing analysis strategies either ignore the spatial component of gene expression variation, or require discretization of the cells into coarse grained groups.\n\nTo address this, we have developed SpatialDE, a computational framework for identifying and characterizing spatially variable genes. Our method generalizes variable gene selection, as used in population-and single-cell studies, to spatial expression profiles. To illustrate the broad utility of our approach, we apply SpatialDE to spatial transcriptomics data, and to data from single cell methods based on multiplexed in situ hybridisation (SeqFISH and MERFISH). SpatialDE enables the statistically robust identification of spatially variable genes, thereby identifying genes with known disease implications, several of which are missed by conventional variable gene selection. Additionally, to enable gene-expressed based histology, SpatialDE implements a spatial gene clustering model which we call \"automatic expression histology,\" allowing to classify genes into groups with distinct spatial patterns.

genomics↗