Biology subjects

Way, G. P.

Publications and source records attributed to Way, G. P..

At least 19 recordsLinked to original sources

bioRxiv2026

On doubting image quality assessment metrics for microscopy virtual staining

Pairing label-free microscopy with virtual staining could reduce the cost and experimental burden of fluorescence microscopy, but its impact is conditional on generalizable inference. Most virtual staining studies assess performance using image quality assessment (IQA) metrics developed for natural images, yet how well these metrics translate to microscopy remains unknown. Here, we examined the behavior of seven commonly-used full-reference training objectives and metrics, MAE, PSNR, SSIM, foreground PSNR and SSIM, LPIPS, and DISTS, under controlled image degradation and realistic out-of-distribution virtual staining. We applied graded intensity, textural, and morphological transformations to Cell Painting images spanning 18 cell lines, seeding densities, and fluorescence channels. Channel, cell line identity and seeding density explained substantial metric variation after controlling for degradation magnitude. DISTS and foreground metrics showed more favorable balance between degradation sensitivity and biological invariance, although no metric reported performance independent of biological context. Incrementally degrading images and evaluating concomitant metric degradation further revealed that most metrics used only a small fraction of their nominal numerical ranges and frequently plateaued while image degradation visibly continued. We next trained three popular virtual staining model architectures (UNet, WGAN-GP, UNeXt) on five U2-OS seeding densities separately, and computed metrics on model predictions across 17 unseen cell lines. We observed that architecture and training U2-OS seeding density together explain less than 2% of metric variation. Visual inspection suggested comparable scores across cell lines correspond to qualitatively distinct errors, such as differences in cell morphology and marker intensity. These findings show that conventional IQA metrics do not effectively translate to virtual staining applications. Selection or optimization of virtual staining models against real application such as in label-free high content drug screening should instead be approached in an application-oriented fashion.

bioinformatics↗

bioRxiv2026

Single-cell hit calling in high-content imaging screens with Buscar

High-content screening (HCS) enables the systematic quantification of single-cell morphology features across thousands of perturbations, capturing rich phenotypic heterogeneity. Image-based profiling is a critical bioinformatics processing step in this pipeline, as researchers use it to predict mechanisms of action, assess toxicity, perform hit calling, and more. However, current image-based profiling workflows rely on aggregate statistics, such as calculating mean or median feature values per well, implicitly assuming cell homogeneity. This limitation obscures subpopulation effects, reducing sensitivity to subtle or heterogeneous effects of perturbations. Here we present Buscar, a method that leverages the full heterogeneity of single-cell image-based profiles to call hits. Buscar requires two reference, single-cell populations that define distinct morphology states: a reference state (e.g., disease cells) and a target state (e.g., healthy cells). Buscar then compares these two groups to define on- and off-morphology signatures, which it then uses to score every perturbation in a given screen. The scores quantify perturbation efficacy and off-target effects, or specificity, in an interpretable manner, clarifying which morphologies are appropriately altered and which may arise from off-target activity. We apply Buscar to three datasets. First, as a proof of concept, we applied Buscar to a Cell Painting dataset of cardiac fibroblasts from patients with heart failure. Buscar quantifies both morphology rescue and off-target morphology activity in these cells treated with a TGF{beta} receptor inhibitor. Second, we show that Buscar recovers biologically coherent gene-phenotype associations across 16 manually-labeled phenotypes in the MitoCheck dataset. Lastly, applied to CPJUMP1, we show that Buscar is robust to technical replicates collected across plates in both small-molecule and CRISPR-Cas9 perturbations. Together, these results establish Buscar as a reproducible and interpretable hit calling method that overcomes aggregation bias, enabling the simultaneous quantification of compound efficacy and specificity to enhance hit calling in HCS. We release Buscar as an open-source python package.

bioinformatics↗

bioRxiv2025

Characterizing the landscape of gene process dependencies in cancer

Precision oncology aims to tailor cancer treatment to tumor genetics but it currently benefits only a small fraction of patients, in part because the primary focus is to match drug targets to single genes. The Cancer Dependency Map Project (DepMap) aimed to characterize the landscape of single-gene dependencies, which increased the universe of potential drug targets. However, the common challenges of drug off-target effects and polypharmacology may limit effectiveness of single genes as drug targets. To address this limitation, we apply BioBombe, an AI/ML framework, to DepMap gene dependency data. This approach characterizes gene process dependencies, which are groups of genes within a biological process that cells rely on for survival. BioBombe fits many hundreds of dimensionality reduction models, across a large range of latent dimensionalities. We find that this multiple-model approach discovers many more gene process dependencies than any single model alone. Using Reactome and CORUM-based gene set enrichment analyses, we characterize the landscape of gene process dependencies, identifying, for example, mitotic regulation and the citric acid cycle as targets, as well as many cancer type-specific dependencies. In gliomas, for example, TP53- and mitochondrial-related pathways emerged as key process vulnerabilities. Linking gene process dependencies with drug sensitivity scores on matched cell lines, we discovered both established and novel drug candidates. Furthermore, we developed a machine learning approach to predict gene process dependencies and associated drug sensitivities from RNA-seq data. Using this approach, we predicted cladribine as a potential new therapeutic for pediatric high-grade glioma and validated its effectiveness at killing these cells. Taken together, BioBombe provides a scalable and interpretable framework for uncovering complex gene process dependencies, which guides drug repurposing, and introduces a novel targeting paradigm for precision oncology.

cancer biology↗

bioRxiv2025

High-content live-cell time-lapse imaging predicts cells about to die via apoptosis

Cell death is a dynamic process that unfolds through time. Live-cell time-lapse imaging captures these dynamics in a way thats impossible for static snapshots. High-content imaging (HCI), which has been developed for static microscopy, applied to time-lapse imaging can quantify how single-cell states change through time. Here we show the ability of high-content live-cell time-lapse imaging (HCLTI) to quantify the onset and progression of one form of cell death called apoptosis. We apply the Live Cell Painting assay called ChromaLIVETM and develop an HCLTI analysis pipeline. We show that HCLTI can discern the morphology dynamics of cells undergoing apoptosis, and demonstrate that machine learning can predict apoptosis as early as 100 minutes after exposing HeLa cells to the apoptosis inducer Staurosporine. This technical advancement paves the way for future studies to better understand the dynamics of other forms of cell death. Understanding cell death dynamics is one piece of solving larger biomedical puzzles like understanding how cells resist death (e.g., therapeutic resistance of cancer cells) and how cells die too soon (e.g., neurodegeneration).

systems biology↗

bioRxiv2025

Stellar quality control for single-cell image-based profiling with coSMicQC

Over the past twenty years, high-content imaging has transformed our ability to measure cell phenotypes. The need to bioinformatically process these phenotypes led to the development of a research field called image-based profiling. However, because the standard image-based profiling approach involves averaging data, single-cell quality control (QC) has been historically ignored. The conventional approach of aggregating single cells into bulk profiles conserves computational resources and reduces, but does not completely remove, the impact of low-quality single cells. As software scalability improves, researchers are increasingly turning to single-cell image-based profiling to reveal important signals of phenotypic heterogeneity. Therefore, this evolution toward single cells compels single-cell QC standards to ensure that observed morphology differences are driven by biology and not technical interference. We address these challenges with coSMicQC (Single cell Morphology Quality Control), a reproducible Python package with comprehensive tutorials that supports systematic, human-in-the-loop filtering of low-quality single cells. CoSMicQC integrates seamlessly into standard image-based profiling protocols, providing an interactive, Jupyter-compatible user interface to set thresholds and flag technical outliers. Applied to four real-world datasets, coSMicQC achieves data-quality gains comparable to labor-intensive manual annotation, at a fraction of the effort. We show how coSMicQC optimizes assay conditions, outperforms the alternative outlier detector PyOD at single-cell phenotype classification, detects mycoplasma contamination, and rescues lead compounds in a large-scale drug screen that would otherwise have been missed. Overall, coSMicQC is a reliable, scalable method for removing technical outliers, reducing noise, and strengthening image-based profiling insights.

bioinformatics↗

bioRxiv2025

Scalable data harmonization for single-cell image-based profiling with CytoTable

High-content imaging (HCI) involves the automated acquisition and quantitative analysis of cell phenotypes from microscopy images. These studies often rely on screening, which can involve thousands of chemical or genetic perturbations that produce terabytes of microscopy data. To extract meaningful biological insights, this data must be processed into quantitative features through a technique known as image-based profiling. A major analytical bottleneck is curating the high-dimensional, single-cell data derived from varied image analysis tools. These datasets suffer from inconsistent schemas, inefficient file formats, and undocumented ontological relationships. These challenges reduce reproducibility and slow progress in downstream applications. To solve these issues, we introduce CytoTable1, a software package for harmonizing single-cell image-based profiling. CytoTable enables modular, portable, and cross-language data integration through a robust, reproducible, and scalable engine that harmonizes single-cell readouts from multiple image analysis tools, preparing for feature integration with software in the Cytomining ecosystem such as Pycytominer.2

bioinformatics↗

bioRxiv2024

Combining NeuroPainting with transcriptomics reveals cell-type-specific morphological and molecular signatures of the 22q11.2 deletion

Neuropsychiatric conditions pose substantial challenges for therapeutic development due to their complex and poorly understood underlying mechanisms. High-throughput, unbiased phenotypic assays present a promising path for advancing therapeutic discovery, especially within disease-relevant neural tissues. Here, we introduce NeuroPainting, a novel adaptation of the Cell Painting assay, optimized for high-dimensional morphological phenotyping of neural cell types, including neurons, neuronal progenitor cells, and astrocytes derived from human stem cells. Using NeuroPainting, we quantified cell structure and organelle behavior across various brain cell types, creating a public dataset of over 4,000 cellular traits. This extensive dataset not only sets a new benchmark for phenotypic screening in neuropsychiatric research but also serves as a gold standard for the research community, enabling comparisons and validation of results. We then applied NeuroPainting to identify morphological signatures associated with the 22q11.2 deletion, a major genetic risk factor for schizophrenia. We observed profound cell-type-specific effects of the 22q11.2 deletion, with significant alterations in mitochondrial structure, endoplasmic reticulum organization, and cytoskeletal dynamics, particularly in astrocytes. Transcriptomic analysis revealed reduced expression of cell adhesion genes in 22q11.2 deletion astrocytes, consistent with recent post-mortem findings. Integrating the RNA sequencing data and morphological profiles uncovered a novel biological link between altered expression of specific cell adhesion molecules and observed changes in mitochondrial morphology in 22q11.2 deletion astrocytes. These findings underscore the power of combined phenomic and transcriptomic analyses to reveal mechanistic insights associated with human genetic variants of neuropsychiatric conditions.

neuroscience↗

bioRxiv2024

High-content microscopy and machine learning characterize a cell morphology signature of NF1 genotype in Schwann cells

Neurofibromatosis type 1 (NF1) is a multi-system, autosomal dominant genetic disorder driven by the systemic loss of the NF1 protein neurofibromin. Loss of neurofibromin in Schwann cells is particularly detrimental, as the acquisition of a second-hit (e.g., complete loss of NF1) can lead to the development of plexiform neurofibromas (pNF). pNFs are painful, disfiguring tumors with an approximately 1 in 5 chance of sarcoma transition. Selumetinib and mirdametinib are currently the only medicines approved by the U.S. Food and Drug Administration (FDA) for the treatment of pNFs. This motivates the need to develop new therapies, either derived to treat NF1 haploinsufficiency or complete loss of NF1 function. To identify new therapies, we need to understand the impact neurofibromin has on Schwann cells. Here, we aimed to characterize differences in high-content microscopy in neurofibromin-deficient Schwann cells. We applied a fluorescence microscopy assay (called Cell Painting) to an isogenic pair of Schwann cell lines (derived from ipn02.3 2{lambda}), one of wildtype genotype (NF1+/+) and one of NF1 null genotype (NF1-/-). We modified the canonical Cell Painting assay to mark four organelles/subcellular compartments: nuclei, endoplasmic reticulum, mitochondria, and F-actin. We utilized CellProfiler to perform quality control, illumination correction, segmentation, and cell morphology feature extraction. We segmented 20,680 NF1 wildtype and null cells, measured 894 significant cell morphology features representing various organelle shapes and intensity patterns, and trained a logistic regression machine learning model to predict the NF1 genotype of single Schwann cells. The machine learning model had high performance, with training and testing data yielding a balanced accuracy of 0.85 and 0.80, respectively. However, when applied to a new pair of Schwann cells, the models balanced accuracy dropped to 0.5, which is no better than random chance. This performance decline appears to result from morphology differences introduced by non-biological factors (cloning procedures, origin of parental cell line, and CRISPR procedures) of the second cell line pair. We plan to improve upon this preliminary model by refining the NF1 morphology signature using a broader panel of Schwann cell lines. Our goal is to apply this enhanced signature in large-scale drug screens of NF1-deficient cells to identify candidate therapeutic agents that specifically reverse the disease-associated morphology. Ultimately, we aim to identify agents that restore NF1 patient-derived Schwann cells to a phenotype resembling the NF1 wild-type and healthier state.

systems biology↗

bioRxiv2024

A morphology and secretome map of pyroptosis

Pyroptosis represents one type of Programmed Cell Death (PCD). It is a form of inflammatory cell death that is canonically defined by caspase-1 cleavage and Gasdermin-mediated membrane pore formation. Caspase-1 initiates the inflammatory response (through IL-1{beta} processing), and the N-terminal cleaved fragment of Gasdermin D polymerizes at the cell periphery forming pores to secrete pro-inflammatory markers. Cell morphology also changes in pyroptosis, with nuclear condensation and membrane rupture. However, recent research challenges canon, revealing a more complex secretome and morphological response in pyroptosis, including overlapping molecular characterization with other forms of cell death, such as apoptosis. Here, we take a multimodal, systems biology approach to characterize pyroptosis. We treated human Peripheral Blood Mononuclear Cells (PBMCs) with 36 different combinations of stimuli to induce pyroptosis or apoptosis. We applied both secretome profiling (nELISA) and high-content fluorescence microscopy (Cell Painting). To differentiate apoptotic, pyroptotic and control cells, we used canonical secretome markers and modified our Cell Painting assay to mark the N-terminus of Gasdermin-D. We trained hundreds of machine learning (ML) models to reveal intricate morphology signatures of pyroptosis that implicate changes across many different organelles and predict levels of many pro-inflammatory markers. Overall, our analysis provides a detailed map of pyroptosis which includes overlapping and distinct connections with apoptosis revealed through a mechanistic link between cell morphology and cell secretome.

systems biology↗

bioRxiv2024

A versatile information retrieval framework for evaluating profile strength and similarity

In profiling assays, thousands of biological properties are measured across many samples, yielding biological discoveries by capturing the state of a cell population, often at the single-cell level. However, for profiling datasets, it has been challenging to evaluate the phenotypic activity of a sample and the phenotypic consistency among samples, due to profiles high dimensionality, heterogeneous nature, and non-linear properties. Existing methods leave researchers uncertain where to draw boundaries between meaningful biological response and technical noise. Here, we developed a statistical framework that uses the well-established mean average precision (mAP) as a single, data-driven metric to bridge this gap. We validated the mAP framework against established metrics through simulations and real-world data applications, revealing its ability to capture subtle and meaningful biological differences in cell state. Specifically, we used mAP to assess both phenotypic activity for a given perturbation (or a sample) as well as consistency within groups of perturbations (or samples) across diverse high-dimensional datasets. We evaluated the framework on different profile types (image, protein, and mRNA profiles), perturbation types (CRISPR gene editing, gene overexpression, and small molecules), and profile resolutions (single-cell and bulk). Our open-source software allows this framework to be applied to identify interesting biological phenomena and promising therapeutics from large-scale profiling data.

bioinformatics↗

bioRxiv2024

Toward generalizable phenotype prediction from single-cell morphology representations

Functional cell processes (e.g., molecular signaling, response to environmental stimuli, mitosis, etc.) impact cell phenotypes, which scientists can easily and robustly measure with cell morphology. However, linking these morphology measurements with phenotypes remains challenging because biologically interpretable phenotypes require manually annotated labels. Automatic phenotype annotation from cell morphology would link biological processes with their phenotypic outcomes and deepen understanding of cell function. We propose that nuclear morphology can be a predictive marker for cell phenotypes that is generalizable across cell types. Nucleus morphology is commonly and easily accessible with microscopy, but annotating specific phenotypic information requires labels. Therefore, we reanalyzed a pre-labeled, publicly-available nucleus microscopy dataset from the MitoCheck consortium to predict single-cell phenotypes. We extracted single-cell morphology features using CellProfiler and DeepProfiler, which provide fast, robust, and generalizable data processing pipelines. We trained multinomial, multi-class elastic net logistic regression models to classify nuclei into one of 15 phenotypes such as Anaphase, Apoptosis, and Binuclear. In a held-out test set, we observed an overall F1 score of 0.84, where individual phenotype scores ranged from 0.64 (indicating moderate performance) to 0.99 (indicating high performance). Notably, phenotypes such as Elongated, Metaphase, and Apoptosis showed high performance. While CellProfiler and DeepProfiler morphology features were generally equally effective, combining feature spaces yielded the best results for 9 of the 15 phenotypes. However, leave-one-image-out (LOIO) cross-validation analysis showed a significant performance decline, indicating our model could not reliably predict phenotype in new single images. Poor performance, which we show was unrelated to factors like illumination correction or model selection, limits generalizability to new datasets and highlights the challenges of morphology to phenotype annotation. Nevertheless, we modified and applied our approach to the JUMP Cell Painting pilot data. Our modified approach improved dataset alignment and highlighted many perturbations that are known to be associated with specific phenotypes. We propose several strategies that could pave the way for more generalizable methods in single-cell phenotype prediction, which is a step toward morphology representation ontologies that would aid in cross-dataset interpretability.

cell biology↗

bioRxiv2023

Molecular subtypes of high grade serous ovarian cancer across racial groups and gene expression platforms

IntroductionHigh-grade serous carcinoma (HGSC) gene expression subtypes are associated with differential survival. We characterized HGSC gene expression in Black individuals and considered whether gene expression differences by race may contribute to poorer HGSC survival among Black versus non-Hispanic White individuals. MethodsWe included newly generated RNA-Seq data from Black and White individuals, and array-based genotyping data from four existing studies of White and Japanese individuals. We assigned subtypes using K-means clustering. Cluster- and dataset-specific gene expression patterns were summarized by moderated t-scores. We compared cluster-specific gene expression patterns across datasets by calculating the correlation between the summarized vectors of moderated t-scores. Following mapping to The Cancer Genome Atlas (TCGA)-derived HGSC subtypes, we used Cox proportional hazards models to estimate subtype-specific survival by dataset. ResultsCluster-specific gene expression was similar across gene expression platforms. Comparing the Black study population to the White and Japanese study populations, the immunoreactive subtype was more common (39% versus 23%-28%) and the differentiated subtype less common (7% versus 22%-31%). Patterns of subtype-specific survival were similar between the Black and White populations with RNA-Seq data; compared to mesenchymal cases, the risk of death was similar for proliferative and differentiated cases and suggestively lower for immunoreactive cases (Black population HR=0.79 [0.55, 1.13], White population HR=0.86 [0.62, 1.19]). ConclusionsA single, platform-agnostic pipeline can be used to assign HGSC gene expression subtypes. While the observed prevalence of HGSC subtypes varied by race, subtype-specific survival was similar. Statement of SignificanceA single pipeline was used to subtype ovarian high-grade serous carcinoma (HGSC) with array-based or RNA-Seq gene expression data. Subtype distributions differed by race, but subtype-specific survival was similar across racial groups.

cancer biology↗

bioRxiv2023

A genome-wide atlas of human cell morphology

A key challenge of the modern genomics era is developing data-driven representations of gene function. Here, we present the first unbiased morphology-based genome-wide perturbation atlas in human cells, containing three genome-scale genotype-phenotype maps comprising >20,000 single-gene CRISPR-Cas9-based knockout experiments in >30 million cells. Our optical pooled cell profiling approach (PERISCOPE) combines a de-stainable high-dimensional phenotyping panel (based on Cell Painting1,2) with optical sequencing of molecular barcodes and a scalable open-source analysis pipeline to facilitate massively parallel screening of pooled perturbation libraries. This approach provides high-dimensional phenotypic profiles of individual cells, while simultaneously enabling interrogation of subcellular processes. Our atlas reconstructs known pathways and protein-protein interaction networks, identifies culture media-specific responses to gene knockout, and clusters thousands of human genes by phenotypic similarity. Using this atlas, we identify the poorly-characterized disease-associated transmembrane protein TMEM251/LYSET as a Golgi-resident protein essential for mannose-6-phosphate-dependent trafficking of lysosomal enzymes, showing the power of these representations. In sum, our atlas and screening technology represent a rich and accessible resource for connecting genes to cellular functions at scale.

genomics↗

bioRxiv2023

High-content microscopy reveals a morphological signature of bortezomib resistance

Drug resistance is a challenge in anticancer therapy, particularly with targeted therapeutics and cytotoxic compounds. In many cases, cancers can be resistant to the drug prior to exposure, i.e., possess intrinsic drug resistance. However, we lack target-independent methods to anticipate resistance in cancer cell lines or characterize intrinsic drug resistance without a priori knowledge of its cause. We hypothesized that cell morphology could provide an unbiased readout of drug sensitivity prior to treatment. We therefore isolated clonal cell lines that were either sensitive or resistant to bortezomib, a well-characterized proteasome inhibitor and anticancer drug to which many cancer cells possess intrinsic resistance. We then measured high-dimensional single-cell morphology profiles using Cell Painting, a high-content microscopy assay. Our imaging- and computation-based profiling pipeline identified morphological features typically different between resistant and sensitive clones. These features were compiled to generate a morphological signature of bortezomib resistance, which correctly predicted the bortezomib treatment response in seven of ten cell lines not included in the training dataset. This signature of resistance was specific to bortezomib over other drugs targeting the ubiquitin-proteasome system. Our results provide evidence that intrinsic morphological features of drug resistance exist and establish a framework for their identification.

cancer biology↗

bioRxiv2023

High-dimensional phenotyping to define the genetic basis of cellular morphology

The morphology of cells is dynamic and mediated by genetic and environmental factors. Characterizing how genetic variation impacts cell morphology can provide an important link between disease association and cellular function. Here, we combined genomic and high-content imaging approaches on iPSCs from 297 unique donors to investigate the relationship between genetic variants and cellular morphology to map what we term cell morphological quantitative trait loci (cmQTLs). We identified novel associations between rare protein altering variants in WASF2, TSPAN15, and PRLR with several morphological traits related to cell shape, nucleic granularity, and mitochondrial distribution. Knockdown of these genes by CRISPRi confirmed their role in cell morphology. Analysis of common variants yielded one significant association and nominated over 300 variants with suggestive evidence (P<10-6) of association with one or more morphology traits. Our results showed that, similar to other molecular phenotypes, morphological profiling can yield insight about the function of genes and variants.

genomics↗

bioRxiv2021

Morphology and gene expression profiling provide complementary information for mapping cell state

Morphological and gene expression profiling can cost-effectively capture thousands of features in thousands of samples across perturbations by disease, mutation, or drug treatments, but it is unclear to what extent the two modalities capture overlapping versus complementary information. Here, using both the L1000 and Cell Painting assays to profile gene expression and cell morphology, respectively, we perturb A549 lung cancer cells with 1,327 small molecules from the Drug Repurposing Hub across six doses, providing a data resource including dose-response data from both assays. The two assays capture both shared and complementary information for mapping cell state. Cell Painting profiles from compound perturbations are more reproducible and show more diversity, but measure fewer distinct groups of features. Applying unsupervised and supervised methods to predict compound mechanisms of action (MOA) and gene targets, we find that the two assays provide a partially shared, but also a complementary view of drug mechanisms. Given the numerous applications of profiling in biology, our analyses provide guidance for planning experiments that profile cells for detecting distinct cell types, disease phenotypes, and response to chemical or genetic perturbations.

bioinformatics↗

bioRxiv2021

Predicting drug polypharmacology from cell morphology readouts using variational autoencoder latent space arithmetic

A variational autoencoder (VAE) is a machine learning algorithm, useful for generating a compressed and interpretable latent space. These representations have been generated from various biomedical data types and can be used to produce realistic-looking simulated data. However, standard vanilla VAEs suffer from entangled and uninformative latent spaces, which can be mitigated using other types of VAEs such as {beta}-VAE and MMD-VAE. In this project, we evaluated the ability of VAEs to learn cell morphology characteristics derived from cell images. We trained and evaluated these three VAE variants--Vanilla VAE, {beta}-VAE, and MMD-VAE--on cell morphology readouts and explored the generative capacity of each model to predict compound polypharmacology (the interactions of a drug with more than one target) using an approach called latent space arithmetic (LSA). To test the generalizability of the strategy, we also trained these VAEs using gene expression data of the same compound perturbations and found that gene expression provides complementary information. We found that the {beta}-VAE and MMD-VAE disentangle morphology signals and reveal a more interpretable latent space. We reliably simulated morphology and gene expression readouts from certain compounds thereby predicting cell states perturbed with compounds of known polypharmacology. Inferring cell state for specific drug mechanisms could aid researchers in developing and identifying targeted therapeutics and categorizing off-target effects in the future.

pharmacology and toxicology↗

bioRxiv2021

Discovering cellular programs of intrinsic and extrinsic drivers of metabolic traits using LipocyteProfiler

A primary obstacle in translating genetics and genomics data into therapeutic strategies is elucidating the cellular programs affected by genetic variants and genes associated with human diseases. Broadly applicable high-throughput, unbiased assays offer a path to rapidly characterize gene and variant function and thus illuminate disease mechanisms. Here, we report LipocyteProfiler, an unbiased high-throughput, high-content microscopy assay that is amenable to large-scale morphological and cellular profiling of lipid-accumulating cell types. We apply LipocyteProfiler to adipocytes and hepatocytes and demonstrate its ability to survey diverse cellular mechanisms by generating rich context-, and process-specific morphological and cellular profiles. We then use LipocyteProfiler to identify known and novel cellular programs altered by polygenic risk of metabolic disease, including insulin resistance, waist-to-hip ratio and the polygenic contribution to lipodystrophy. LipocyteProfiler paves the way for large-scale forward and reverse phenotypic profiling in lipid-storing cells, and provides a framework for the unbiased identification of causal relationships between genetic variants and cellular programs relevant to human disease.

genomics↗