Evaluating batch correction methods for image-based cell profiling

Nature Communications, Год журнала: 2024, Номер 15(1)

Опубликована: Авг. 2, 2024

Abstract High-throughput image-based profiling platforms are powerful technologies capable of collecting data from billions cells exposed to thousands perturbations in a time- and cost-effective manner. Therefore, has been increasingly used for diverse biological applications, such as predicting drug mechanism action or gene function. However, batch effects severely limit community-wide efforts integrate interpret collected across different laboratories equipment. To address this problem, we benchmark ten high-performing single-cell RNA sequencing (scRNA-seq) correction techniques, representing approaches, using newly released Cell Painting dataset, JUMP. We focus on five scenarios with varying complexity, ranging batches prepared single lab over time imaged microscopes multiple labs. find that Harmony Seurat RPCA noteworthy, consistently ranking among the top three methods all tested while maintaining computational efficiency. Our proposed framework, benchmark, metrics can be assess new future. This work paves way improvements enable community make best use public scientific discovery.

Язык: Английский

---

Exploring NLRP3-related phenotypic fingerprints in human macrophages using Cell Painting assay

iScience, Год журнала: 2025, Номер 28(3), С. 111961 - 111961

Опубликована: Фев. 6, 2025

Existing research has proven difficult to understand the interplay between upstream signaling events during NLRP3 inflammasome activation. Additionally, downstream of complex formation such as cytokine release and pyroptosis can exhibit variation, further complicating matters. Cell Painting emerged a prominent tool for unbiased evaluation effect perturbations on cell morphological phenotypes. Using this technique, phenotypic fingerprints be generated that reveal connections phenotypes possible modes action. To best our knowledge, was first study utilized human THP-1 macrophages generate in response different endogenous exogenous triggers identify features specific formation. Our results demonstrated not only are trigger-specific but it also cellular associated with

Язык: Английский

---

Capturing cell heterogeneity in representations of cell populations for image-based profiling using contrastive learning

PLoS Computational Biology, Год журнала: 2024, Номер 20(11), С. e1012547 - e1012547

Опубликована: Ноя. 11, 2024

Image-based cell profiling is a powerful tool that compares perturbed populations by measuring thousands of single-cell features and summarizing them into profiles. Typically sample represented averaging across cells, but this fails to capture the heterogeneity within populations. We introduce CytoSummaryNet: Deep Sets-based approach improves mechanism action prediction 30-68% in mean average precision compared on public dataset. CytoSummaryNet uses self-supervised contrastive learning multiple-instance framework, providing an easier-to-apply method for aggregating feature data than previously published strategies. Interpretability analysis suggests model achieves improvement downweighting small mitotic cells or those with debris prioritizing large uncrowded cells. The requires only perturbation labels training, which are readily available all datasets. offers straightforward post-processing step profiles can significantly boost retrieval performance image-based

Язык: Английский

---

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

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Ноя. 17, 2024

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

Язык: Английский

---

Capturing cell heterogeneity in representations of cell populations for image-based profiling using contrastive learning

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2023, Номер unknown

Опубликована: Ноя. 16, 2023

Image-based cell profiling is a powerful tool that compares perturbed populations by measuring thousands of single-cell features and summarizing them into profiles. Typically sample represented averaging across cells, but this fails to capture the heterogeneity within populations. We introduce CytoSummaryNet: Deep Sets-based approach improves mechanism action prediction 30-68% in mean average precision compared on public dataset. CytoSummaryNet uses self-supervised contrastive learning multiple-instance framework, providing an easier-to-apply method for aggregating feature data than previously published strategies. Interpretability analysis suggests model achieves improvement downweighting small mitotic cells or those with debris prioritizing large uncrowded cells. The requires only perturbation labels training, which are readily available all datasets. offers straightforward post-processing step profiles can significantly boost retrieval performance image-based

Язык: Английский

---

Cell morphological representations of genes enhance prediction of drug targets

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Июнь 10, 2024

Abstract Identifying how a given chemical of interest exerts its impact on biological systems is critical step in developing new medicines and products. The mechanism query compound can sometimes be identified when image-based morphological profile matches library well-annotated profiles. In this study, we demonstrate significant improvement classification performance by incorporating side information: gene representations. We generate these representations using the profiles cells where level single gene’s expression has been artificially increased or decreased. genes are selected as those encoding known protein targets annotated compounds library. A transformer model trained to classify gene-compound pairs, each pair represents potential interaction between compound, true false. Subsequently, generates ranked list likely target for previously unseen compound. Although strategy exhibits high only that encountered – due limited size our training dataset increase demonstrates notable over simply matching directly Larger datasets may improve prediction capabilities approach, enabling novel compounds, which then experimentally validated.

Язык: Английский

---

Cell morphology and gene expression: tracking changes and complementarity across time and cell lines

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Сен. 1, 2024

Summary Effective drug discovery relies on combining target knowledge with functional assays and multi-omics data to understand chemicals’ molecular actions. However, the relationship between cell morphology gene expression over time across lines remains unclear. To explore this, we analyzed Cell Painting L1000 for 106 compounds three from osteoblast, lung, breast tumors (U2OS, A549, MCF7) at points (6h, 24h, 48h) using a 10µM concentration. We found significant line effects in data, less pronounced transcriptomics. Using Weighted Gene Co-expression Network Analysis (WGCNA) enrichment analysis, identified connections deregulation chemicals similar biological (e.g., HDAC CDK inhibitors). These findings suggest that while shows distinct patterns, both technologies offer complementary insights into compound-induced cellular changes, enhancing chemical risk assessment.

Язык: Английский

---

A scalable, data analytics workflow for image-based morphological profiles

Chemometrics and Intelligent Laboratory Systems, Год журнала: 2024, Номер unknown, С. 105232 - 105232

Опубликована: Сен. 1, 2024

Язык: Английский

---

Insights into the Identification of iPSC- and Monocyte-Derived Macrophage-Polarizing Compounds by AI-Fueled Cell Painting Analysis Tools

International Journal of Molecular Sciences, Год журнала: 2024, Номер 25(22), С. 12330 - 12330

Опубликована: Ноя. 17, 2024

Macrophage polarization critically contributes to a multitude of human pathologies. Hence, modulating macrophage is promising approach with enormous therapeutic potential. Macrophages are characterized by remarkable functional and phenotypic plasticity, pro-inflammatory (M1) anti-inflammatory (M2) states at the extremes multidimensional spectrum. Cell morphology major indicator for activation, describing M1(-like) (rounded) M2(-like) (elongated) different cell shapes. Here, we introduced painting macrophages better reflect their multifaceted plasticity associated phenotypes beyond rigid dichotomous M1/M2 classification. Using high-content imaging, established deep learning- feature-based image analysis tools elucidate cellular fingerprints that inform about subtle blood monocyte-derived iPSC-derived as screening surrogate. Moreover, show feature profiling suitable identifying inter-donor variance describe relevance 'cell roundness' dissect distinct signatures after stimulation known biological or small-molecule modulators (re-)polarization. Our novel AI-fueled provide resource high-content-based drug candidate profiling, which set stage (re-)polarization in health disease.

Язык: Английский

---

Developing and Applying Computational Models to Uncover Mechanistic Insights into Complex Biological Processes Across Molecular, Cellular, and Systemic Levels

Next frontier., Год журнала: 2024, Номер 8(1), С. 173 - 173

Опубликована: Ноя. 25, 2024

The complexity of biological processes spans molecular, cellular, and systemic levels, requiring advanced computational models to unravel the intricate mechanisms underlying these phenomena. This research explores development application gain mechanistic insights into diverse systems. By integrating multi-scale data from genomics, proteomics, cellular imaging, this study leverages machine learning algorithms, dynamical systems modeling, network analysis simulate analyze interactions. Key areas focus include understanding signaling pathways, differentiation, physiological responses. also highlights role tools in bridging experimental with theoretical predictions, providing a robust framework for hypothesis generation testing. Challenges such as heterogeneity, scalability, model interpretability are addressed, emphasizing need interdisciplinary approaches. aims advance field biology by offering novel complex fostering applications personalized medicine, drug development, synthetic biology.

Язык: Английский

---