Biology subjects

Bar-Joseph, Z.

Publications and source records attributed to Bar-Joseph, Z..

At least 19 recordsLinked to original sources

bioRxiv2026

Foundation Models Improve Perturbation Response Prediction

Predicting cellular responses to genetic or chemical perturbations has been a long-standing goal in biology. Recent applications of foundation models to this task have yielded contradictory results regarding their superiority over simple baselines. We conducted an extensive analysis of over 600 different models across various prediction tasks and evaluation metrics, demonstrating that while some foundation models fail to outperform simple baselines, others significantly improve predictions for both genetic and chemical perturbations. Furthermore, we developed and evaluated methods for integrating multiple foundation models for perturbation prediction. Our results show that with sufficient data, these models approach fundamental performance limits, confirming that foundation models can improve cellular response simulations. Code and Data: https://github.com/genbio-ai/foundation-models-perturbation

bioinformatics↗

bioRxiv2025

Finetuning Foundation Models for Temporal Clinical Transcriptomics Data

BackgroundTimeseries clinical transcriptomic datasets offer the opportunity to gain insights into the dynamics of disease mechanisms/treatment responses. However, their utility in uncovering temporal patterns is often limited by high noise levels and small sample sizes. Leveraging foundational gene embeddings and incorporating interaction information can help address these challenges, improve gene network analysis, and enable the detection of subtle changes that drive disease progression or drug response. ResultsWe finetuned gene embeddings from foundation models using healthy tissue gene expression data and used them in temporal GNNs to model gene expression of responder and non-responders to treatment in 3 disease datasets - ulcerative colitis, Crohns disease and psoriasis. Application of our method to these datasets confirmed known mechanisms associated with drug action, and also identified key differences between activated and repressed pathways for responders and non responders including B-Cell activation and mitochondria related activity in ulcerative colitis patients. ConclusionFinetuning gene embeddings from foundation models provide a richer context to model gene expression data compared to using them in their naive state. Even with smaller sample sizes, results from GNN-based temporal models outperform traditional methods by detecting known mechanisms of response and unraveling role of genes and mechanisms not known to be associated with response and non-response. Code AvailabilityCode and data are available in a public GitHub repository - https://github.com/Sanofi-Public/GNN-Timeseries

bioinformatics↗

bioRxiv2025

An Intrinsic-hoc Framework for Heterogeneous Cellular Senescence Elucidation Using Deep Graph Representation Learning and Experimental Validation

Cellular senescence is a primordial driver of tissue and organ aging, and the accumulation of senescent cells (SnCs) has been implicated in numerous age-related diseases. A major barrier to studying senescence is the rarity and heterogeneity of SnCs, which are not a uniform population but instead comprise diverse senotypes shaped by cell-of-origin and microenvironmental context. Such heterogeneity exceeds what classical senescence hallmarks can resolve at single-cell resolution, motivating the need for computational frameworks that can capture senotype-level diversity intrinsically. Here, we introduce DeepSAS, a deep graph representation learning framework that robustly identifies cell-type-specific SnCs and their senescence-associated genes (SnGs). DeepSAS incorporates a heterogeneous graph that integrates intracellular transcriptional states with intercellular communication cues, enabling the joint inference of senescent cells and senescence-linked genes through attention-based contrastive learning. Applied to public healthy eye and lung atlases, DeepSAS identified SnCs whose proportions positively correlate with aging. From in-house idiopathic pulmonary fibrosis (IPF) patient scRNA-seq data, DeepSAS detected 1,678 SnCs (out of 24,125 cells) and 263 SnGs across 26 cell types, including 43 SnGs that are uniquely associated with a single cell type. We generated high-resolution Xenium spatial transcriptomics data to further validate SnGs in IPF, revealing NFE2L2 as a SnG specifically enriched in CTHRC1+ fibroblasts. Notably, the ex vivo bleomycin-induced senescence in human precision-cut lung slice (hPCLS) samples similarly identified NFE2L2 as an SnG in CTHRC1+ fibroblasts, albeit with stronger transcriptional signals, suggesting mechanistic differences in senescence cells associated with chronic and acute injury. Overall, DeepSAS uncovers distinct senescence programs and infers cell-type-specific SnGs that are difficult to resolve using existing marker-based approaches. We believe it offers a generalizable and translationally relevant strategy for advancing senescence biology and therapeutic development.

bioinformatics↗

bioRxiv2025

Active Learning for Protein Structure Prediction

Accurate protein structure prediction is challenging especially for families and types that are not well-represented in current training data. Active learning selects candidates for labeling with the aim of most rapidly improving model performance. In a general, many labeling strategies have been proposed. However, with protein structure prediction, most of these strategies dont apply or are difficult due to the high-dimensional, variable-dimensional regression target and the inherent complexity of the models involved. We applied a novel active learning strategy, DEW-DROP, to protein structure prediction on two different protein datasets: VHH-only antibodies (Nanobodies), and Mycobacterial proteins. We introduce a domain-specific fine-tuned Equifold model for VHH structures and apply DEWDROP to generate ensembles of predictions using Monte Carlo dropout. Using the statistics of these we select batches with high information content for labeling. We show that DEWDROP (1) improves model training efficiency through batch optimization outper-forming baselines, and (2) selects data with relevant high information content.

molecular biology↗

bioRxiv2024

SenSet, a novel human lung senescence cell gene signature, identifies cell-specific senescence mechanisms

Cellular senescence is a major hallmark of aging. Senescence is defined as an irreversible growth arrest observed when cells are exposed to a variety of stressors including DNA damage, oxidative stress, or nutrient deprivation. While senescence is a well-established driver of aging and age-related diseases, it is a highly heterogeneous process with significant variations across organisms, tissues, and cell types. The relatively low abundance of senescence in healthy aged tissues represents a major challenge to studying senescence in a given organ, including the human lung. To overcome this limitation, we developed a Positive-Unlabeled (PU) learning framework to generate a comprehensive senescence marker gene list in human lungs (termed SenSet) using the largest publicly available single-cell lung dataset, the Human Lung Cell Atlas (HLCA). We validated SenSet in a highly complex ex vivo human 3D lung tissue culture model subjected to the senescence inducers bleomycin, doxorubicin, or irradiation, and established its sensitivity and accuracy in characterizing senescence. Using SenSet, we identified and validated cell-type specific senescence signatures in distinct lung cell populations upon aging and environmental exposures. Our study presents the first comprehensive analysis of senescent cells in the healthy aging lung and uncovers cell-specific gene signatures of senescence, presenting fundamental implications for our understanding of major lung diseases, including cancer, fibrosis, chronic obstructive pulmonary disease, or asthma.

genomics↗

bioRxiv2024

ENACT: End-to-End Analysis of Visium High Definition (HD) Data

MotivationSpatial transcriptomics (ST) enables the study of gene expression within its spatial context in histopathology samples. To date, a limiting factor has been the resolution of sequencing based ST products. The introduction of the Visium High Definition (HD) technology opens the door to cell resolution ST studies. However, challenges remain in the ability to accurately map transcripts to cells and in assigning cell types based on the transcript data. ResultsWe developed ENACT, the first tissue-agnostic pipeline that integrates advanced cell segmentation with Visium HD transcriptomics data to infer cell types across whole tissue sections. Our pipeline incorporates novel bin-to-cell assignment methods, enhancing the accuracy of single-cell transcript estimates. Validated on diverse synthetic and real datasets, our approach is both scalable and effective offering a robust solution for spatially resolved transcriptomics analysis. Availability and implementationENACT source code is available at https://github.com/Sanofi-Public/enact-pipeline. Experimental data is available at https://zenodo.org/records/13887921. Supplementary information: Supplementary data are available at BiorXiv online.

bioinformatics↗

bioRxiv2024

Detecting m6A RNA modification from nanopore sequencing using a semi-supervised learning framework

Direct nanopore-based RNA sequencing can be used to detect post-transcriptional base modifications, such as m6A methylation, based on the electric current signals produced by the distinct chemical structures of modified bases. A key challenge is the scarcity of adequate training data with known methylation modifications. We present Xron, a hybrid encoder-decoder framework that delivers a direct methylation-distinguishing basecaller by training on synthetic RNA data and immunoprecipitation-based experimental data in two steps. First, we generate data with more diverse modification combinations through in silico cross-linking. Second, we use this dataset to train an end-to-end neural network basecaller followed by fine-tuning on immunoprecipitation-based experimental data with label-smoothing. The trained neural network basecaller outperforms existing methylation detection methods on both read-level and site-level prediction scores. Xron is a standalone, end-to-end m6A-distinguishing basecaller capable of detecting methylated bases directly from raw sequencing signals, enabling de novo methylome assembly.

bioinformatics↗

bioRxiv2023

scResolve: Recovering single cell expression profiles from multi-cellular spatial transcriptomics

Many popular spatial transcriptomics techniques lack single-cell resolution. Instead, these methods measure the collective gene expression for each location from a mixture of cells, potentially containing multiple cell types. Here, we developed scResolve, a method for recovering single-cell expression profiles from spatial transcriptomics measurements at multi-cellular resolution. scResolve accurately restores expression profiles of individual cells at their locations, which is unattainable from cell type deconvolution. Applications of scResolve on human breast cancer data and human lung disease data demonstrate that scResolve enables cell type-specific differential gene expression analysis between different tissue contexts and accurate identification of rare cell populations. The spatially resolved cellular-level expression profiles obtained through scResolve facilitate more flexible and precise spatial analysis that complements raw multi-cellular level analysis.

bioinformatics↗

bioRxiv2023

CodonBERT: Large Language Models for mRNA Design and Optimization

AO_SCPLOWBSTRACTC_SCPLOWmRNA based vaccines and therapeutics are gaining popularity and usage across a wide range of conditions. One of the critical issues when designing such mRNAs is sequence optimization. Even small proteins or peptides can be encoded by an enormously large number of mRNAs. The actual mRNA sequence can have a large impact on several properties including expression, stability, immunogenicity, and more. To enable the selection of an optimal sequence, we developed CodonBERT, a large language model (LLM) for mRNAs. Unlike prior models, CodonBERT uses codons as inputs which enables it to learn better representations. CodonBERT was trained using more than 10 million mRNA sequences from a diverse set of organisms. The resulting model captures important biological concepts. CodonBERT can also be extended to perform prediction tasks for various mRNA properties. CodonBERT outperforms previous mRNA prediction methods including on a new flu vaccine dataset.

bioinformatics↗

bioRxiv2023

Optimal transport for mapping senescent cells in spatial transcriptomics

Spatial transcriptomics (ST) provides a unique opportunity to study cellular organization and cell-cell interactions at the molecular level. However, due to the low resolution of the sequencing data additional information is required to utilize this technology, especially for cases where only a few cells are present for important cell types. To enable the use of ST to study senescence we developed scDOT, which combines ST and single cell RNA-Sequencing (scRNA-Seq) to improve the ability to reconstruct single cell resolved spatial maps. scDOT integrates optimal transport and expression deconvolution to learn non-linear couplings between cells and spots and to infer cell placements. Application of scDOT to existing and new lung ST data improves on prior methods and allows the identification of the spatial organization of senescent cells, the identification of their neighboring cells and the identification of novel genes involved in cell-cell interactions that may be driving senescence.

cell biology↗

bioRxiv2023

Deep Batch Active Learning for Drug Discovery

A key challenge in drug discovery is to optimize, in silico, various absorption and affinity properties of small molecules. One strategy that was proposed for such optimization process is active learning. In active learning molecules are selected for testing based on their likelihood of improving model performance. To enable the use of active learning with advanced neural network models we developed two novel active learning batch selection methods. These methods were tested on several public datasets for different optimization goals and with different sizes. We have also curated new affinity datasets that provide chronological information on state-of-the-art experimental strategy. As we show, for all datasets the new active learning methods greatly improved on existing and current batch selection methods leading to significant potential saving in the number of experiments needed to reach the same model performance. Our methods are general and can be used with any package including the popular DeepChem library.

bioinformatics↗

bioRxiv2023

Modelling Human Post-Implantation Development via Extra-Embryonic Niche Engineering

Implantation of the human embryo commences a critical developmental stage that comprises profound morphogenetic alteration of embryonic and extra-embryonic tissues, axis formation, and gastrulation events. Our mechanistic knowledge of this window of human life remains limited due to restricted access to in vivo samples for both technical and ethical reasons. Additionally, human stem cell models of early post-implantation development with both embryonic and extra-embryonic tissue morphogenesis are lacking. Here, we present iDiscoid, produced from human induced pluripotent stem cells via an engineered a synthetic gene circuit. iDiscoids exhibit reciprocal co-development of human embryonic tissue and engineered extra-embryonic niche in a model of human post-implantation. They exhibit unanticipated self-organization and tissue boundary formation that recapitulates yolk sac-like tissue specification with extra-embryonic mesoderm and hematopoietic characteristics, the formation of bilaminar disc-like embryonic morphology, the development of an amniotic-like cavity, and acquisition of an anterior-like hypoblast pole and posterior-like axis. iDiscoids offer an easy-to-use, high-throughput, reproducible, and scalable platform to probe multifaceted aspects of human early post-implantation development. Thus, they have the potential to provide a tractable human model for drug testing, developmental toxicology, and disease modeling.

developmental biology↗

bioRxiv2023

Transcription Factor Dynamics in Cross-Regulation of Plant Hormone Signaling Pathways

Cross-regulation between hormone signaling pathways is indispensable for plant growth and development. However, the molecular mechanisms by which multiple hormones interact and co-ordinate activity need to be understood. Here, we generated a cross-regulation network explaining how hormone signals are integrated from multiple pathways in etiolated Arabidopsis (Arabidopsis thaliana) seedlings. To do so we comprehensively characterized transcription factor activity during plant hormone responses and reconstructed dynamic transcriptional regulatory models for six hormones; abscisic acid, brassinosteroid, ethylene, jasmonic acid, salicylic acid and strigolactone/karrikin. These models incorporated target data for hundreds of transcription factors and thousands of protein-protein interactions. Each hormone recruited different combinations of transcription factors, a subset of which were shared between hormones. Hub target genes existed within hormone transcriptional networks, exhibiting transcription factor activity themselves. In addition, a group of MITOGEN-ACTIVATED PROTEIN KINASES (MPKs) were identified as potential key points of cross-regulation between multiple hormones. Accordingly, the loss of function of one of these (MPK6) disrupted the global proteome, phosphoproteome and transcriptome during hormone responses. Lastly, we determined that all hormones drive substantial alternative splicing that has distinct effects on the transcriptome compared with differential gene expression, acting in early hormone responses. These results provide a comprehensive understanding of the common features of plant transcriptional regulatory pathways and how cross-regulation between hormones acts upon gene expression.

plant biology↗

bioRxiv2023

Cell segmentation for high-resolution spatial transcriptomics

Spatial transcriptomics promises to greatly improve our understanding of tissue organization and cell-cell interactions. While most current platforms for spatial transcriptomics only offer multi-cellular resolution, with 10-15 cells per spot, recent technologies provide a much denser spot placement leading to sub-cellular resolution. A key challenge for these newer methods is cell segmentation and the assignment of spots to cells. Traditional image-based segmentation methods are limited and do not make full use of the information profiled by spatial transcrip-tomics. Here we present SCS, which combines imaging data with sequencing data to improve cell segmentation accuracy. SCS assigns spots to cells by adaptively learning the position of each spot relative to the center of its cell using a transformer neural network. SCS was tested on two new sub-cellular spatial transcriptomics technologies and outperformed traditional image-based segmentation methods. SCS achieved better accuracy, identified more cells, and provided more realistic cell size estimation. Sub-cellular analysis of RNAs using SCS spots assignments provides information on RNA localization and further supports the segmentation results.

bioinformatics↗

bioRxiv2022

A unified analysis of atlas single cell data

Recent efforts to generate atlas-scale single cell data provide opportunities for joint analysis across tissues and across modalities. Most of the existing methods for single cell atlas analysis use cells as the reference unit to combine datasets. However, such methods suffer from the inability to effectively integrate cross-modality data, hindering downstream gene-based analysis, and loss of genuine biological variations. Here we present a new data integration method, GIANT, which is for the first time designed for the atlas-scale analysis from the gene perspective. GIANT first converts datasets from different modalities into gene graphs, and then recursively embeds genes in the graphs into a latent space without additional alignment. Applying GIANT to the HuBMAP datasets creates a unified gene embedding space across multiple human tissues and data modalities, where gene representations reflect the functions of genes in their cells. Further evaluations demonstrate the usefulness of GIANT in discovering diverse gene functions, and underlying gene regulations in cells of different tissues.

bioinformatics↗

bioRxiv2021

Unsupervised cell functional annotation for single-cell RNA-Seq

One of the first steps in the analysis of single cell RNA-Sequencing data (scRNA-Seq) is the assignment of cell types. While a number of supervised methods have been developed for this, in most cases such assignment is performed by first clustering cells in low-dimensional space and then assigning cell types to different clusters. To overcome noise and to improve cell type assignments we developed UNIFAN, a neural network method that simultaneously clusters and annotates cells using known gene sets. UNIFAN combines both, low-dimensional representation for all genes and cell specific gene set activity scores to determine the clustering. We applied UNIFAN to human and mouse scRNA-Seq datasets from several different organs. As we show, by using knowledge on gene sets, UNIFAN greatly outperforms prior methods developed for clustering scRNA-Seq data. The gene sets assigned by UNIFAN to different clusters provide strong evidence for the cell type that is represented by this cluster making annotations easier. Softwarehttps://github.com/doraadong/UNIFAN

bioinformatics↗

bioRxiv2021

Inferring cell-cell interactions from pseudotime ordering of scRNA-Seq data

A major advantage of single cell RNA-Sequencing (scRNA-Seq) data is the ability to reconstruct continuous ordering and trajectories for cells. To date, such ordering was mainly used to group cells and to infer interactions within cells. Here we present TraSig, a computational method for improving the inference of cell-cell interactions in scRNA-Seq studies. Unlike prior methods that only focus on the average expression levels of genes in clusters or cell types, TraSig fully utilizes the dynamic information to identify significant ligand-receptor pairs with similar trajectories, which in turn are used to score interacting cell clusters. We applied TraSig to several scRNA-Seq datasets. As we show, using the ordering information allows TraSig to obtain unique predictions that improve upon those identified by prior methods. Functional experiments validate the ability of TraSig to identify novel signaling interactions that impact vascular development in liver organoid.

bioinformatics↗

bioRxiv2021

Cellar: Interactive single cell data annotation tool

Several recent technologies and platforms enable the profiling of various molecular signals at the single-cell level. A key question for all studies using such data is the assignment of cell types. To improve the ability to correctly assign cell types in single and multi-omics sequencing and imaging single-cell studies, we developed Cellar. This interactive software tool supports all steps in the analysis and assignment process. We demonstrate the advantages of Cellar by using it to annotate several HuBMAP datasets from multi-omics single-cell sequencing and spatial proteomics studies. Cellar is freely available and includes several annotated reference HuBMAP datasets. Availabilityhttps://data.test.hubmapconsortium.org/app/cellar

bioinformatics↗