Capturing single-cell heterogeneity via data fusion improves image-based profiling

Rohban, Mohammad Hossein; Abbasi, Hamdah Shafqat; Singh, Shantanu; Carpenter, Anne E.

doi:10.1038/s41467-019-10154-8

Cited by 33 publications

(32 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Alternatively, it may reflect an implicit encoding of the heterogeneity of single cells within a microscopy image 134 . Capturing single-cell heterogeneity after feature extraction has indeed been found to improve image-based phenotypic clustering 161 . Finally, the flexible architecture of neural networks enables information to flow in from alternative data sources and formats, as input, or as side information.…”

Section: Future Directionsmentioning

confidence: 99%

Image-based profiling for drug discovery: due for a machine-learning upgrade?

Chandrasekaran

Ceulemans

Boyd

et al. 2020

Nat Rev Drug Discov

Self Cite

288

242

View full text Add to dashboard Cite

Image-based profiling is a maturing strategy by which the rich information present in biological images is reduced to a multidimensional profile, a collection of extracted image-based features. These profiles can be mined for relevant patterns, revealing unexpected biological activity that is useful for many steps in the drug discovery process. Such applications include identifying disease-associated screenable phenotypes, understanding disease mechanisms and predicting a drug’s activity, toxicity or mechanism of action. Several of these applications have been recently validated and have moved into production mode within academia and the pharmaceutical industry. Some of these have yielded disappointing results in practice but are now of renewed interest due to improved machine-learning strategies that better leverage image-based information. Although challenges remain, novel computational technologies such as deep learning and single-cell methods that better capture the biological information in images hold promise for accelerating drug discovery.

show abstract

Section: Future Directionsmentioning

confidence: 99%

Image-based profiling for drug discovery: due for a machine-learning upgrade?

Chandrasekaran

Ceulemans

Boyd

et al. 2020

Nat Rev Drug Discov

Self Cite

288

242

View full text Add to dashboard Cite

show abstract

“…Beyond VEGF and eNOS expression, our data contained information on cell morphology, actin structure and cell adjacency that we had not yet utilized. Based on previous work that found a spatial correlation in the proliferation of PAEC (Gien et al, 2007), work showing that morphology can be used as a predictor of cellular response to molecular intervention, and recent work suggesting that genotypic and phenotypic traits are passed from a mother cell to daughter cells (Phillips et al, 2019; Rohban et al, 2019; Shaffer et al, 2018, preprint; Singh et al, 2015) we hypothesized that we would find spatial clustering of cells with similar overall responses to IGF-1 administration.…”

Section: Resultsmentioning

confidence: 99%

Heterogeneous response of endothelial cells to insulin like growth factor 1 treatment is explained by spatially clustered subpopulations

2019

View full text Add to dashboard Cite

A common strategy to measure the efficacy of drug treatment is the in vitro comparison of ensemble readouts with and without treatment, such as proliferation and cell death. A fundamental assumption underlying this approach is that there exists minimal cell-to-cell variability in the response to a drug. Here, we demonstrate that ensemble and non-spatial single-cell readouts applied to primary cells may lead to incomplete conclusions due to cell-to-cell variability. We exposed primary fetal pulmonary artery endothelial cells (PAEC) isolated from healthy newborn sheep and persistent pulmonary hypertension of the newborn (PPHN) sheep to the growth hormone, insulin-like growth factor 1 (IGF-1). We found that IGF-1 increased proliferation and branch points in tube formation assays but not angiogenic signaling proteins at the population level for both cell types. We hypothesized that this molecular ambiguity was due to the presence of cellular sub-populations with variable responses to IGF-1. Using high throughput single-cell imaging, we discovered a spatially localized response to IGF-1. This suggests localized signaling or heritable cell response to external stimuli may ultimately be responsible for our observations. Discovering and further exploring these rare cells is critical to finding new molecular targets to restore cellular function.

show abstract

“…Recent advances in single CTC analysis provide new ways to collect critical phenotypic information, which reveals the tumor heterogeneity among CTCs population. [ 18,19 ] As a basic unit of life, individual cells provide better resolution of biology than bulk population, which offers new opportunity to analyze the expression level on single CTC resolution. [ 20 ] Trapping or patterning of individual cells without the interference of intercellular interactions (crosstalk) is fundamental for single cell analysis.…”

Section: Introductionmentioning

confidence: 99%

Self‐Assembled GO Honeycomb Microarray for Selective Cancer Cell Capture and Single Cell Analysis of Proteolytic Expression

Chen

2020

Adv Healthcare Materials

View full text Add to dashboard Cite

Proteolytic enzymes expressed by circulating tumor cells are proved to facilitate their invasion into multiple organs via cleaving natural ECM networks, leading to consequent metastasis colonization and malignant lethality. Recent evidence suggests the rare metastasis initiating cells with higher proteolytic levels among circulating tumor cells (CTCs) may strongly increase the risk of metastasis. Beyond selective CTC capture, the heterogeneity in proteases expression provides a promising indicator for metastasis happening. To this end, the graphene oxide (GO) honeycomb microarray with single CTC matched sizes is fabricated via the self-assembly breath figure approach, which serves as an integrated protocol for selective CTC capture and single-cell analysis of protease activity. Contributing to synergistic effects of structure and chemistry, CTCs can be efficiently isolated and individually trapped in each honeycomb hole. Meanwhile, the crosstalk among CTCs can be erased by blocking direct cell-to-cell contact, which offers promising potentials in the single-cell analysis of protease expression. Integrating specific capture and in situ analysis of single CTCs on GO micropatterned surface is of significant importance in various biological and clinical applications such as cancer diagnostics and cancer therapeutic evaluation.

show abstract

Capturing single-cell heterogeneity via data fusion improves image-based profiling

Cited by 33 publications

References 19 publications

Image-based profiling for drug discovery: due for a machine-learning upgrade?

Image-based profiling for drug discovery: due for a machine-learning upgrade?

Heterogeneous response of endothelial cells to insulin like growth factor 1 treatment is explained by spatially clustered subpopulations

Self‐Assembled GO Honeycomb Microarray for Selective Cancer Cell Capture and Single Cell Analysis of Proteolytic Expression

Contact Info

Product

Resources

About