Biology subjects

Blair, J. D.

Publications and source records attributed to Blair, J. D..

12 recordsLinked to original sources

bioRxiv2026

Systematic identification of seed-driven off-target effects in Perturb-seq experiments

Genome-wide Perturb-seq (GWPS) has emerged as a powerful approach for unbiased mapping of gene regulatory networks. A key assumption underlying many Perturb-seq analyses is that each guide RNA exclusively perturbs a single target locus. Without methods to identify and filter off-target events, erroneous gene-pathway associations driven by off-target activity can propagate into downstream analyses. Here, we present a workflow for the systematic identification of candidate off-target events in CRISPRi Perturb-seq experiments. Our approach exploits the observation that cells harboring a guide which represses an off-target gene display transcriptional similarity to cells in which that gene is directly targeted by an on-target guide. We apply our workflow to multiple GWPS datasets and nominate off-target events in which a guide nominally targeting one gene also represses a distinct gene producing a phenotype likely attributable to the off-target perturbation. We use both off-target gene repression and guide seed sequence alignments at the off-target promoter locus as evidence for off-target effects and find independent evidence of putative off-target events in separate GWPS datasets. Together, these results establish a principled framework for the identification and filtering of off-target guide effects in Perturb-seq experiments.

genomics↗

bioRxiv2026

Genome-wide single-cell perturbation screens with VIPerturb-seq

CRISPR-based screening combined with single-cell sequencing (i.e., Perturb-seq) enables systematic mapping of genetic perturbations to molecular phenotypes. While Perturb-seq is well-suited to profile targeted subsets of regulators, scaling to genome-wide screens presents substantial cost and throughput challenges. Here we introduce VIPerturb-seq, a platform to facilitate routine genome-wide Perturb-seq experiments using probe-based detection workflows. We describe a split probe strategy for detection of genome-wide CRISPR libraries in fixed cells that enables (i) support for phenotypic enrichment of Very Important Perturbations (VIPs) prior to single-cell profiling, and (ii) compatibility with combinatorial indexing workflows to further improve Perturb-seq throughput by 50-fold. Using a genome-wide CRISPRi library (GuEST-List), we demonstrate VIPerturb-seq on three genome-wide screens representing both unbiased and phenotypically enriched workflows. Our results demonstrate how the sensitivity, scalability, and efficiency of VIPerturb-seq can enable both individual labs with targeted research questions and large data generation platforms aiming to construct virtual cells.

genomics↗

bioRxiv2025

Iterative, multimodal, and scalable single-cell profiling for discovery and characterization of signaling regulators

Cell signaling plays a critical role in regulating cellular state, yet uncovering regulators of signaling pathways and understanding their molecular consequences remains challenging. Here, we present an iterative experimental and computational framework to identify and characterize regulators of signaling proteins, using the mTOR marker phosphorylated RPS6 (pRPS6) as a case study. We present a customized workflow that uses the 10x Flex assay to jointly profile intracellular protein levels, transcriptomes, and CRISPR perturbations in single cells. We use this to generate a "glossary" dataset of paired protein-RNA measurements across targeted perturbations, which we leverage to train a predictive model of pRPS6 levels based solely on transcriptomic data. Applying this model to a genome-wide Perturb-seq dataset enables in silico screening for pRPS6 and nominates novel regulators of mTOR signaling. Experimental validation confirms these predictions and reveals mechanistic diversity among hits, including changes in signaling output driven by anabolic activity, cellular proliferation and multiple stress pathways. Our work demonstrates how integrated experimental and computational approaches provide a scalable framework for multimodal phenotyping and discovery.

genomics↗

bioRxiv2025

Leveraging synthetic data produced from museum specimens to train adaptable species classification models

Computer vision has increasingly shown potential to improve data processing efficiency in ecological research. However, training computer vision models requires large amounts of high-quality, annotated training data. This poses a significant challenge for researchers looking to create bespoke computer vision models, as substantial human resources and biological replicates are often needed to adequately train these models. Synthetic images have been proposed as a potential solution for generating large training datasets, but models trained with synthetic images often have poor generalization to real photographs. Here we present a modular pipeline for training generalizable classification models using synthetic images. Our pipeline includes 3D asset creation with the use of 3D scanners, synthetic image generation with open-source computer graphic software, and domain adaptive classification model training. We demonstrate our pipeline by applying it to skulls of 16 mammal species in the order Carnivora. We explore several domain adaptation techniques, including maximum mean discrepancy (MMD) loss, fine-tuning, and data supplementation. Using our pipeline, we were able to improve classification accuracy on real photographs from 55.4% to a maximum of 95.1%. We also conducted qualitative analysis with t-distributed stochastic neighbor embedding (t-SNE) and gradient-weighted class activation mapping (Grad-CAM) to compare different domain adaptation techniques. Our results demonstrate the feasibility of using synthetic images for ecological computer vision and highlight the potential of museum specimens and 3D assets for scalable, generalizable model training.

bioinformatics↗

bioRxiv2025

mTORC1 activation drives astrocyte reactivity in cortical tubers and brain organoid models of TSC

Tuberous Sclerosis Complex (TSC) is a genetic neurodevelopmental disorder associated with early onset epilepsy, intellectual disability and neuropsychiatric disorders. A hallmark of the disorder is cortical tubers, which are focal malformations of brain development containing dysplastic cells with hyperactive mTORC1 signaling. One barrier to developing therapeutic approaches and understanding the origins of tuber cells is the lack of a model system that recapitulates this pathology. To address this, we established a genetically mosaic cortical organoid system that models a somatic "second-hit" mutation, which is thought to drive the formation of tubers in TSC. With this model, we find that loss of TSC2 cell-autonomously promotes the differentiation of astrocytes, which exhibit features of a disease-associated reactive state. TSC2-/- astrocytes have pronounced changes in morphology and upregulation of proteins that are risk factors for neurodegenerative diseases, such as clusterin and APOE. Using multiplexed immunofluorescence in primary tubers from TSC patients, we show that tuber cells with hyperactive mTORC1 activity also express reactive astrocyte proteins, and we identify a unique population of cells with expression profiles that match those observed in organoids. Together, this work reveals that reactive astrogliosis is a primary feature of TSC that arises early in cortical development. Dysfunctional glia are therefore poised to be drivers of pathophysiology, nominating a potential therapeutic target for treating TSC and related mTORopathies.

neuroscience↗

bioRxiv2024

A hybrid approach to invertebrate biomonitoring using computer vision and DNA metabarcoding

O_LIAutomated invertebrate classification using computer vision has shown significant potential to improve specimen processing efficiency. However, challenges such as invertebrate diversity and morphological similarity among taxa can make it difficult to infer fine-scale taxonomic classifications using computer vision. As a result, many invertebrate computer vision models are forced to make classifications at coarser levels, such as at family or order. C_LIO_LIHere we propose a novel modular method to combine computer vision and bulk DNA metabarcoding specimen processing pipelines to improve the accuracy and taxonomic granularity of individual specimen classifications. To improve specimen classification accuracy, our methods use multimodal fusion models that combine image data with DNA-based assemblage data. To refine the taxonomic granularity of the models classifications, our methods cross-references the classifications with DNA metabarcoding detections from bulk samples. We demonstrated these methods using a continental-scale, invertebrate bycatch dataset collected by the National Ecological Observatory Network. We also introduce the CV.eDNA R package, which aims to assist practitioners looking to implement our methods. C_LIO_LIUsing our methods, we reached a classification accuracy of 79.6% across the 17 taxa using real DNA assemblage data, and 83.6% when the assemblage data was "error-free", resulting in a 2.2% and 6.2% increase in accuracy when compared to a model trained using only images. After cross-referencing with the DNA metabarcoding detections, we improved taxonomic granularity in up to 72.2% of classifications, with up to 5.7% reaching species-level. C_LIO_LIBy providing computer vision models with coincident DNA assemblage data, and refining individual classifications using DNA metabarcoding detections, our methods the potential to greatly expand the capabilities of biological computer vision classifiers. Our methods allow computer vision classifiers to infer taxonomically fine-grained classifications when it would otherwise be difficult or impossible due to challenges of morphologic similarity or data scarcity. These methods are not limited to terrestrial invertebrates and could be applied in any instance where image and DNA metabarcoding data are concurrently collected. C_LI

bioinformatics↗

bioRxiv2024

Single-cell genotype-phenotype mapping identifies therapeutic vulnerabilities in VEXAS syndrome

Somatic evolution leads to the emergence of clonal diversity across tissues with broad implications for human health. A striking example of somatic evolution is the VEXAS (Vacuoles E1 enzyme X-linked Autoinflammatory Somatic) syndrome, caused by somatic UBA1 mutations in hematopoietic stem cells (HSCs), inducing treatment-refractory, systemic inflammation. However, the mechanisms that lead to survival and expansion of mutant HSCs are unknown, limiting the development of effective therapies. The lack of animal or cellular models of UBA1-mutant HSCs has hindered such mechanistic understanding, mandating analysis of primary human VEXAS samples, which harbor admixtures of wild-type and UBA1-mutant HSCs. To address these challenges, we applied single-cell multi-omics to comprehensively define mutant UBA1-induced transcriptome, chromatin accessibility and signaling pathway alterations in VEXAS patients, allowing for the direct comparison of mutant versus wild-type cells within the same environment. We confirmed the expected enrichment of UBA1M41V/T mutations in myeloid cells, and additionally discovered that these mutations were also prevalent in Natural Killer (NK) cells in VEXAS patients, providing new insights into disease phenotypes. Through mapping genotypes to molecular phenotypes, including transcriptome, chromatin accessibility, cell surface protein or intracellular protein profiles, in HSCs, we found that UBA1M41V/T-mutant cells showed an increased inflammation signature (interferon alpha and gamma response pathways), as well as activation of unfolded protein response (UPR) via pro-survival, but not pro-apoptotic, mediators of the PERK pathway, compared to UBA1 wild-type HSCs. Ex vivo validation experiments showed that inhibiting UBA1 in normal CD34+ or using UBA1-mutant HSCs led to PERK pathway up-regulation, increased myeloid differentiation and cell survival, which was reversed by PERK inhibition. Thus, we demonstrated that human VEXAS HSCs show cell-intrinsic inflammatory phenotypes and survive the proteomic stress caused by compromised ubiquitination through PERK-mediated activation of the UPR. Together, these analyses nominate PERK pathway inhibition as a potential new therapeutic strategy for eradicating the VEXAS-inducing clone, demonstrating the power of single-cell multi-omics profiling of primary patient sample to enable genotype-to-phenotype somatic mapping for the discovery of novel candidates for clinical intervention.

genomics↗

bioRxiv2023

STAT3 regulates the generation of astroglia inhuman brain organoids with high mTORC1 activity

During brain development, neural progenitor cells first produce neurons, then astrocytes and other glial cell types, which provide important trophic support and shape neuronal development and function. Intrinsic genetic programs interact with extracellular signals to control progenitor fate, resulting in temporally segregated periods of neurogenesis and gliogenesis. Animal models have implicated STAT3 as an important driver of astrogenesis; however, the signaling pathways that control glial differentiation during human brain development are less well understood. Prior work demonstrated that constitutive activation of mTORC1 signaling in human brain organoid models resulted in the precocious generation of glial-lineage cells. In this study, we tested whether mTORC1 acts via STAT3 to control astrogenesis in brain organoids. We show that knockdown of STAT3 reduces astrogenesis in wild-type organoids and in organoids with constitutively high mTORC1 signaling caused by deletion of the negative regulator TSC2. However, mTORC1 is not required for cytokine-induced activation of STAT3 and expression of the astrocytic protein GFAP. Together, these results show that mTORC1 acts through STAT3 to control astroglia production in human brain organoid models, but that mTOR signaling is dispensable for STAT3-driven astrogenesis. Summary statementDeb et al, use human brain organoid models to show a requirement for STAT3 downstream of mTORC1 in regulating astrogliogenesis during early human brain development.

neuroscience↗

bioRxiv2023

Phospho-seq: Integrated, multi-modal profiling of intracellular protein dynamics in single cells

Cell signaling plays a critical role in regulating cellular behavior and fate. While multimodal single-cell sequencing technologies are rapidly advancing, scalable and flexible profiling of cell signaling states alongside other molecular modalities remains challenging. Here we present Phospho-seq, an integrated approach that aims to quantify phosphorylated intracellular and intranuclear proteins, and to connect their activity with cis-regulatory elements and transcriptional targets. We utilize a simplified benchtop antibody conjugation method to create large custom antibody panels for simultaneous protein and scATAC-seq profiling on whole cells, and integrate this information with scRNA-seq datasets via bridge integration. We apply our workflow to cell lines, induced pluripotent stem cells, and 3-month-old brain organoids to demonstrate its broad applicability. We demonstrate that Phospho-seq can define cellular states and trajectories, reconstruct gene regulatory relationships, and characterize the causes and consequences of heterogeneous cell signaling in neurodevelopment.

genomics↗

bioRxiv2021

Allochronic isolation on a latitudinal gradient in the polyphagous moth Hyphantria cunea

O_LIAllochronic speciation, where reproductive isolation between populations of a species is facilitated by a difference in reproductive timing, depends on abiotic factors such as seasonality and biotic factors such as diapause intensity. These factors are strongly influenced by latitudinal trends in climate, so we hypothesized that there is a relationship between latitude and divergence among populations separated by life history timing. C_LIO_LIHyphantria cunea (the fall webworm), a Lepidopteran defoliator with red and black colour morphs, is hypothesized to be undergoing an incipient allochronic speciation. However, given their broad geographic range, the strength of allochronic speciation may vary across latitude. We annotated >11,000 crowd-sourced observations of fall webworm to model geographic distribution, phenology, and differences in colour phenotype between morphs across North America. C_LIO_LIWe found that red and black morph life history timing differs across North America, and the phenology of morphs diverges more in warmer climates at lower latitudes. We also found some evidence that the colour phenotype of morphs also diverges at lower latitudes, suggesting reduced gene flow between colour morphs. C_LIO_LIOur results demonstrate that seasonality in lower latitudes may increase the strength of allochronic speciation in insects, and that the strength of sympatric speciation can vary along a latitudinal gradient. This has implications for our understanding of broad-scale speciation events and trends in global biodiversity. C_LI

evolutionary biology↗

bioRxiv2021

Embracing imperfection: machine-assisted invertebrate classification in real-world datasets

O_LIDespite growing concerns over the health of global invertebrate diversity, terrestrial invertebrate monitoring efforts remain poorly geographically distributed. Machine-assisted classification has been proposed as a potential solution to quickly gather large amounts of data; however, previous studies have often used unrealistic or idealized datasets to train their models. C_LIO_LIIn this study, we describe a practical methodology for including machine learning in ecological data acquisition pipelines. Here we train and test machine learning algorithms to classify over 56,000 bulk terrestrial invertebrate specimens from morphometric data and contextual metadata. All vouchered specimens were collected in pitfall traps by the National Ecological Observatory Network (NEON) at 27 locations across the United States in 2016. Specimens were photographed, and morphometric data was extracted as feature vectors using ImageJ. Issues stemming from inconsistent taxonomic label specificity were resolved by making classifications at the lowest identified taxonomic level (LITL). Taxa with too few specimens to be included in the training dataset were classified by the model using zero-shot classification. C_LIO_LIWhen classifying specimens that were known and seen by our models, we reached an accuracy of 72.7% using extreme gradient boosting (XGBoost) at the LITL. Models that were trained without contextual metadata underperformed models with contextual metadata by an average of 7.2%. We also classified invertebrate taxa that were unknown to the model using zero-shot classification, with an accuracy of 39.4%, resulting in an overall accuracy of 71.5% across the entire NEON dataset. C_LIO_LIThe general methodology outlined here represents a realistic application of machine learning as a tool for ecological studies. Hierarchical and LITL classifications allow for flexible taxonomic specificity at the input and output layers. These methods also help address the long tail problem of underrepresented taxa missed by machine learning models. Finally, we encourage researchers to consider more than just morphometric data when training their models, as we have shown that the inclusion of contextual metadata can provide significant improvements to accuracy. C_LI

ecology↗

bioRxiv2017

Widespread translational remodeling during human neuronal differentiation

Faithful cellular differentiation requires precise coordination of changes in gene expression. However, the relative contributions of transcriptional and translational regulation during human cellular differentiation are unclear. Here, we induced forebrain neuronal differentiation of human embryonic stem cells (hESCs) and characterized genomewide RNA and translation levels during neurogenesis. We find that thousands of genes change at the translation level across differentiation without a corresponding change in RNA level. Specifically, we identify mTOR complex 1 signaling as a key driver for elevated translation of translation-related genes in hESCs. In contrast, translational repression in active neurons is mediated by transcript 3' UTRs, through regulatory sequences. Together, our findings identify a functional role for the dramatic 3' UTR extensions that occur during brain development, and provide insights to interpret genetic variants in post-transcriptional control factors that influence neurodevelopmental disorders and diseases.

cell biology↗