Ihuan Gunawan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Ihuan Gunawan has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- John George Lock
School of Biomedical Sciences, University of New South Wales
6 shared publications - Erik Meijering
School of Computer Science and Engineering, University of New South Wales, Sydney, Australia
4 shared publications - Daniel P. Neumann
School of Biomedical Sciences, Faculty of Medicine and Health, University of New South Wales, Sydney, NSW, Australia
3 shared publications - Fatemeh Vafaee
School of Biotechnology and Biomolecular Sciences, University of New South Wales, Sydney, NSW 2052, Australia
3 shared publications - Felix V. Kohane
School of Biomedical Sciences, Faculty of Medicine and Health, University of New South Wales, Sydney, New South Wales, 2052, Australia
3 shared publications - Nandini Aggarwal2 shared publications
Latest Publications
- Open Access
Generative semantic multiplexing (SemaPlex) for accessible and scalable multiplexed fluorescence imaging
Multiplexed fluorescence imaging enhances spatially-resolved interrogation of complex, multi-molecular cell processes that are insufficiently sampled using standard 4-5 plex imaging. To improve accessibility and scalability for multiplexed imaging, we demonstrate generative ‘Semantic Multiplexing’ (SemaPlex); a simple experimental and deep learning strategy for amplifying marker plexity several-fold by semantically unmixing multiple markers combined per imaging channel. We first characterise key determinants of SemaPlex performance, achieving precise computational multiplexing of 2-to-8 markers synthetically mixed in one channel, facilitating enhanced cell phenotype classification. We then demonstrate practical SemaPlex application, acquiring 10 markers over 4 channels (3*3-plex+1) to efficiently emulate real multiplexed labelling. This permitted accurate reconstruction of quantitative single-cell phenotypic manifolds delineating cell-cycle and mitotic dynamics, with internally validated error-detection. Finally, we exemplify use of ‘semantic guides’; additional input channels that significantly enhance multiplexing fidelity. SemaPlex makes multiple-fold increases in fluorescence imaging-plexity accessible, scalable and customisable; democratising multiplexed imaging-based interrogation of complex cell biology.
Open AccessGenerative semantic multiplexing (SemaPlex) for accessible and scalable multiplexed fluorescence imaging
Multiplexed fluorescence imaging enhances spatially-resolved interrogation of complex, multi-molecular cell processes that are insufficiently sampled using standard 4-5 plex imaging. To improve accessibility and scalability for multiplexed imaging, we demonstrate generative ‘Semantic Multiplexing’ (SemaPlex); a simple experimental and deep learning strategy for amplifying marker plexity several-fold by semantically unmixing multiple markers combined per imaging channel. We first characterise key determinants of SemaPlex performance, achieving precise computational multiplexing of 2-to-8 markers synthetically mixed in one channel, facilitating enhanced cell phenotype classification. We then demonstrate practical SemaPlex application, acquiring 10 markers over 4 channels (3*3-plex+1) to efficiently emulate real multiplexed labelling. This permitted accurate reconstruction of quantitative single-cell phenotypic manifolds delineating cell-cycle and mitotic dynamics, with internally validated error-detection. Finally, we exemplify use of ‘semantic guides’; additional input channels that significantly enhance multiplexing fidelity. SemaPlex makes multiple-fold increases in fluorescence imaging-plexity accessible, scalable and customisable; democratising multiplexed imaging-based interrogation of complex cell biology.
- Open Access
Image quality metrics fail to accurately represent biological information in fluorescence microscopy
Image processing methods offer the potential to improve the quality of fluorescence microscopy data, allowing for image acquisition at lower, less phototoxic illumination doses. The training and evaluation of such methods is informed and driven by full-reference image quality metrics (IQMs); however, these metrics derive from applications to natural scene images, not fluorescence microscopy images. Here we investigate the response of IQMs to common properties of fluorescence microscopy data and whether IQMs are capable of reporting the biological information content of images. We find that IQM scores are biased by image content for both raw and processed microscopy data, and that improvements in IQM values reported after processing are not reliably correlated with performance in downstream analysis tasks. As common IQMs are unreliable proxies for guiding image processing developments in biological fluorescence microscopy, image processing performance should be benchmarked according to downstream analysis success.
Open AccessImage quality metrics fail to accurately represent biological information in fluorescence microscopy
Image processing methods offer the potential to improve the quality of fluorescence microscopy data, allowing for image acquisition at lower, less phototoxic illumination doses. The training and evaluation of such methods is informed and driven by full-reference image quality metrics (IQMs); however, these metrics derive from applications to natural scene images, not fluorescence microscopy images. Here we investigate the response of IQMs to common properties of fluorescence microscopy data and whether IQMs are capable of reporting the biological information content of images. We find that IQM scores are biased by image content for both raw and processed microscopy data, and that improvements in IQM values reported after processing are not reliably correlated with performance in downstream analysis tasks. As common IQMs are unreliable proxies for guiding image processing developments in biological fluorescence microscopy, image processing performance should be benchmarked according to downstream analysis success.
- 5 August 2025
- Published by EMBL-EBI
The effect of image acquisition settings and image processing on fluorescence microscopy data and downstream analysis
BioStudies – one package for all the data supporting a study
- 5 August 2025
- Published by EMBL-EBI
The effect of image acquisition settings and image processing on fluorescence microscopy data and downstream analysis
BioStudies – one package for all the data supporting a study
- Open Access
Epithelial–mesenchymal cell state heterogeneity predetermines differential phospho-signaling responses to EGF stimulation
Understanding why isogenic cancer cells respond differently to equivalent oncogenic stimuli is vital for optimizing anticancer therapies. Emerging evidence suggests that pre-existing differences in cell state may modulate signaling responses to new stimuli, but the interplay of specific cell states and signals remains unclear. We investigated whether epithelial–mesenchymal (E/M) state, a major axis of cancer cell heterogeneity, influences signaling responses to epidermal growth factor (EGF), a critical oncogenic stimulus in non-small cell lung cancer (NSCLC). We imaged >64,000 A549 NSCLC cells labeled for DNA, F-actin and alternate signaling markers (p-AKT-S473, p-AKT-T308, p-ERK or p-S6) after acute stimulation. Quantitative single-cell morphological and spatial profiling defined a stimulus-invariant ‘E/M state landscape’ over which EGF signaling responses were compared. This revealed state-dependent differences in signal-activation magnitudes, dynamics and subcellular routing. AKT responses exhibited phosphosite- and compartment-specific dynamics across states, with epithelial cells showing strong, transient membrane-localized S473 and higher internalized T308, whereas mesenchymal cells displayed weaker but sustained nuclear and ruffle-localized S473. Regression-based computational multiplexing concurrently inferred all signaling responses per cell, mapping state-dependent divergence in multi-molecular signaling trajectories. E/M state thus pre-determines distinctive spatiotemporal profiles of EGF-induced signaling, with implications for signaling functions and anti-signaling therapy responses across E/M state-diverse tumors.
Open AccessEpithelial–mesenchymal cell state heterogeneity predetermines differential phospho-signaling responses to EGF stimulation
Understanding why isogenic cancer cells respond differently to equivalent oncogenic stimuli is vital for optimizing anticancer therapies. Emerging evidence suggests that pre-existing differences in cell state may modulate signaling responses to new stimuli, but the interplay of specific cell states and signals remains unclear. We investigated whether epithelial–mesenchymal (E/M) state, a major axis of cancer cell heterogeneity, influences signaling responses to epidermal growth factor (EGF), a critical oncogenic stimulus in non-small cell lung cancer (NSCLC). We imaged >64,000 A549 NSCLC cells labeled for DNA, F-actin and alternate signaling markers (p-AKT-S473, p-AKT-T308, p-ERK or p-S6) after acute stimulation. Quantitative single-cell morphological and spatial profiling defined a stimulus-invariant ‘E/M state landscape’ over which EGF signaling responses were compared. This revealed state-dependent differences in signal-activation magnitudes, dynamics and subcellular routing. AKT responses exhibited phosphosite- and compartment-specific dynamics across states, with epithelial cells showing strong, transient membrane-localized S473 and higher internalized T308, whereas mesenchymal cells displayed weaker but sustained nuclear and ruffle-localized S473. Regression-based computational multiplexing concurrently inferred all signaling responses per cell, mapping state-dependent divergence in multi-molecular signaling trajectories. E/M state thus pre-determines distinctive spatiotemporal profiles of EGF-induced signaling, with implications for signaling functions and anti-signaling therapy responses across E/M state-diverse tumors.
- Open Access
- 17 May 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-15
Extensible Immunofluorescence (ExIF) accessibly generates high-plexity datasets by integrating standard 4-plex imaging data
Standard immunofluorescence imaging captures just ~4 molecular markers (4-plex) per cell, limiting dissection of complex biology. Inspired by multimodal omics-based data integration approaches, we propose an Extensible Immunofluorescence (ExIF) framework that transforms carefully designed but easily produced panels of 4-plex immunofluorescence into a unified dataset with theoretically unlimited marker plexity, using generative deep learning-based virtual labelling. ExIF enables integrated analyses of complex cell biology, exemplified here through interrogation of the epithelial-mesenchymal transition (EMT), driving significant improvements in downstream quantitative analyses usually reserved for omics data, including: classification of cell phenotypes; manifold learning of cell phenotype heterogeneity; and pseudotemporal inference of molecular marker dynamics. Introducing data integration concepts from omics to microscopy, ExIF empowers life scientists to use routine 4-plex fluorescence microscopy to quantitatively interrogate complex, multimolecular single-cell processes in a manner that approaches the performance of multiplexed labelling methods whose uptake remains limited.
- 17 May 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-15
Open AccessExtensible Immunofluorescence (ExIF) accessibly generates high-plexity datasets by integrating standard 4-plex imaging data
Standard immunofluorescence imaging captures just ~4 molecular markers (4-plex) per cell, limiting dissection of complex biology. Inspired by multimodal omics-based data integration approaches, we propose an Extensible Immunofluorescence (ExIF) framework that transforms carefully designed but easily produced panels of 4-plex immunofluorescence into a unified dataset with theoretically unlimited marker plexity, using generative deep learning-based virtual labelling. ExIF enables integrated analyses of complex cell biology, exemplified here through interrogation of the epithelial-mesenchymal transition (EMT), driving significant improvements in downstream quantitative analyses usually reserved for omics data, including: classification of cell phenotypes; manifold learning of cell phenotype heterogeneity; and pseudotemporal inference of molecular marker dynamics. Introducing data integration concepts from omics to microscopy, ExIF empowers life scientists to use routine 4-plex fluorescence microscopy to quantitatively interrogate complex, multimolecular single-cell processes in a manner that approaches the performance of multiplexed labelling methods whose uptake remains limited.
- Open Access
- 2 August 2023
- Published by Elsevier in Cell Reports Methods
- Vol. 3 (8) , 100547
An introduction to representation learning for single-cell data analysis
- 2 August 2023
- Published by Elsevier in Cell Reports Methods
- Vol. 3 (8) , 100547
Open AccessAn introduction to representation learning for single-cell data analysis