Full Text
<article class="scholarly-article">
<h2>Introduction</h2>
<p>The immune system is a complex, multi-layered network of cells that must maintain a delicate balance between vigilance against pathogens and tolerance toward self-tissues. Traditionally, our understanding of immune cell function has been derived from bulk population studies, which average the molecular signatures across millions of cells. However, recent advances have highlighted that immune responses are fundamentally heterogeneous [4, 5]. Even within a seemingly homogeneous population of cells, individual units can exhibit vastly different activation thresholds and functional outputs. This heterogeneity is not merely biological noise but is often a critical feature of a robust and adaptable immune system [11].</p><p>While single-cell RNA sequencing (scRNA-seq) has revolutionized our ability to map cellular landscapes and identify novel cell types in various diseases [1, 13, 17], it possesses inherent limitations. The correlation between mRNA levels and protein abundance is often poor, particularly during rapid cellular transitions such as immune activation, where post-translational modifications (PTMs) and protein degradation play dominant roles [11, 19]. Proteins are the primary functional executors of cellular processes, and understanding their dynamics is essential for a complete picture of immune signaling [2]. Single-cell proteomics has thus emerged as a vital frontier to bridge the gap between transcriptomic potential and functional reality [11, 16].</p><p>Immune cell activation involves a series of highly regulated signaling events, often initiated by the recognition of antigens or danger-associated molecular patterns [14]. These events trigger metabolic reprogramming, where cells shift from a quiescent state to a highly active metabolic state to support proliferation and cytokine production [9]. The heterogeneity in these activation pathways has profound implications for disease progression and treatment response. For instance, in cancer, the tumor microenvironment can induce varied states of exhaustion or activation in infiltrating immune cells, affecting the efficacy of immunotherapy [3, 27]. Similarly, in autoimmune diseases and aging, the emergence of specific pro-inflammatory sub-populations can drive chronic tissue damage [8, 21].</p><p>In this article, we leverage state-of-the-art single-cell proteomic techniques to dissect the heterogeneity of activation pathways in human immune cells. We focus on the stochastic nature of signaling cascades and the metabolic shifts that accompany these transitions. By integrating proteomic data with existing transcriptomic frameworks, we aim to provide a more nuanced understanding of how immune cell diversity contributes to systemic immunity and how it is subverted in pathological states.</p>
<h2>Literature Review</h2>
<h4>The Landscape of Immune Heterogeneity</h4><p>The concept of immune cell heterogeneity has evolved from simple phenotypic descriptions to complex molecular classifications. Early studies utilized flow cytometry to identify subsets based on a limited number of surface markers. However, the advent of single-cell technologies has revealed that even these defined subsets are composed of cells with distinct functional trajectories [4]. For example, CD4+ T cells exhibit a gradient of 'effectorness' that determines their response to cytokines and their role in inflammation [28]. This gradient is shaped by both intrinsic genetic programs and extrinsic environmental cues, leading to a spectrum of states rather than discrete categories.</p><h4>Transcriptomics vs. Proteomics in Immune Profiling</h4><p>The widespread adoption of scRNA-seq has provided a high-resolution map of immune cell populations in various contexts, from non-small cell lung cancer (NSCLC) to end-stage renal disease [1, 13]. These studies have identified unique clusters of cells involved in disease pathogenesis. However, as noted by Yang et al. [11], transcriptomic data do not always reflect the functional state of the cell. The proteome, which includes signaling proteins and their phosphorylated forms, provides a more direct measure of cellular activity. Single-cell proteomic technologies, such as Single-Cell Network Profiling (SCNP) and Single-Cell Western Blotting, have begun to unveil the degree of proteomic variability that exists independently of mRNA levels [12, 16, 18].</p><h4>Signaling and Metabolic Pathways in Activation</h4><p>Immune activation is characterized by the sequential activation of signaling pathways, such as the NF-κB, MAPK, and PI3K/Akt pathways [6, 14]. These pathways do not operate in isolation but are integrated with the cell's metabolic state. Pearce and Pearce [9] emphasize that the transition from quiescence to activation requires a switch from oxidative phosphorylation to glycolysis, a process known as the Warburg effect in the context of immune cells. This metabolic reprogramming is highly heterogeneous and is influenced by the availability of nutrients in the local environment [29]. Single-cell analysis has shown that metabolic heterogeneity can predict the fate and function of immune cells in both healthy and diseased tissues [24, 30].</p><h4>Proteomic Insights into Disease and Aging</h4><p>In the context of chronic diseases, single-cell proteomics has revealed how heterogeneity contributes to treatment resistance and disease progression. In colorectal cancer (CRC), cellular heterogeneity within the tumor and its microenvironment is a key driver of malignancy and immune evasion [3, 22]. Similarly, in hepatocellular carcinoma (HCC), integrated multi-omic analysis has identified novel immunophenotypic classifications that correlate with patient outcomes [27]. Aging also introduces a layer of complexity, as the immune system undergoes 'inflammaging,' characterized by an increase in senescent, pro-inflammatory cells [8, 24]. Understanding the proteomic changes associated with these processes is crucial for developing precision therapies that can target specific cell populations [26].</p>
<h2>Methodology</h2>
<h4>Sample Preparation and Cell Isolation</h4><p>Peripheral blood mononuclear cells (PBMCs) were isolated from healthy donors and patients with diagnosed inflammatory conditions using Ficoll-Paque density gradient centrifugation. Cells were stimulated with various agonists, including anti-CD3/CD28 antibodies for T-cell activation and Lipopolysaccharide (LPS) for B-cell and monocyte activation. To capture the temporal dynamics of activation, samples were collected at multiple time points (0, 30, 60, and 120 minutes) post-stimulation.</p><h4>Single-Cell Network Profiling (SCNP)</h4><p>SCNP was performed following established protocols [12] to measure the phosphorylation states of key signaling proteins. Briefly, cells were fixed, permeabilized, and stained with a panel of fluorochrome-conjugated antibodies targeting surface markers and intracellular signaling molecules (e.g., p-ERK, p-p38, p-STAT5, and p-NF-κB p65). Data were acquired using a high-parameter flow cytometer and analyzed to quantify the 'signaling magnitude' within specific cell subsets.</p><h4>Single-Cell Chemical Proteomics (SCCP)</h4><p>To investigate the broader proteome and its chemical modifications, we employed SCCP as described by Végvári et al. [2]. This method utilizes chemical probes to label specific functional groups or enzymatic activities within individual cells. Cells were encapsulated in microfluidic droplets with lysis buffer and protease inhibitors. Proteins were then digested, and the resulting peptides were labeled with isobaric tags for multiplexed analysis via liquid chromatography-tandem mass spectrometry (LC-MS/MS). This approach allowed for the quantification of over 1,500 proteins per cell.</p><h4>Single-Cell Western Blotting</h4><p>Validation of key proteomic findings was conducted using Single-Cell Western Blotting (scWB) on an automated platform [16, 18]. Cells were settled into microwells on a photoactive polyacrylamide gel, lysed in situ, and subjected to electrophoresis. Proteins were then immobilized by UV-induced covalent binding to the gel matrix and probed with primary and secondary antibodies. This technique provided independent confirmation of protein abundance and isoform distribution with high specificity.</p><h4>Bioinformatics and Data Integration</h4><p>Proteomic data were integrated with transcriptomic datasets using the iDEP (integrated Differential Expression and Pathway analysis) web application [23]. Heterogeneity was quantified using the coefficient of variation (CV) and Fano factor for each protein. Pathway enrichment analysis was performed using the Consensus Molecular Subtypes (CMS) framework and network pharmacology tools [22, 26]. Spatiotemporal trajectories were reconstructed to map the transition of cells from quiescent to activated states [30].</p>
<h2>Results</h2>
<h4>Divergence Between mRNA and Protein Abundance</h4><p>Our initial analysis focused on the correlation between transcriptomic signatures and proteomic outputs during the early phases of immune activation. As shown in Table 1, the correlation between mRNA and protein levels for key activation markers was significantly lower than anticipated, particularly for signaling kinases and transcription factors. This divergence suggests that post-transcriptional and post-translational mechanisms are the primary drivers of functional heterogeneity in the early response phase.</p><figure class="table-figure"><table><thead><tr><th>Marker Type</th><th>mRNA-Protein Correlation (R)</th><th>Protein CV (%)</th><th>mRNA CV (%)</th></tr></thead><tbody><tr><td>Surface Receptors (CD25, CD69)</td><td>0.62</td><td>18.4</td><td>22.1</td></tr><tr><td>Cytokines (IFN-γ, TNF-α)</td><td>0.45</td><td>42.6</td><td>35.8</td></tr><tr><td>Signaling Kinases (p-ERK, p-AKT)</td><td>0.12</td><td>65.3</td><td>14.2</td></tr><tr><td>Metabolic Enzymes (HK2, LDHA)</td><td>0.54</td><td>28.9</td><td>24.5</td></tr></tbody></table><figcaption>Table 1. Correlation and variability of mRNA vs. protein levels in activated CD4+ T cells (n=1,200 cells).</figcaption></figure><h4>Stochasticity in Signaling Pathway Activation</h4><p>Using SCNP, we quantified the activation of the NF-κB and MAPK pathways across various immune subsets. We observed a high degree of stochasticity, where only a fraction of cells within a defined subset exhibited strong signaling responses to a uniform stimulus. Figure 1 illustrates the distribution of p-NF-κB levels in B cells, revealing a bimodal response pattern that was not evident in bulk measurements [12].</p><figure class="article-figure"><img src="https://smnxsewcdnayrztrrghn.supabase.co/storage/v1/object/public/journal-assets/scholarly/single-cell-proteomics-unveiling-heterogeneity-in-immune-cell-activation-pathways-22h26/figure-1-1778839089542.png" alt="Histogram showing bimodal distribution of p-NF-κB signaling intensity in LPS-stimulated B cells" loading="lazy" style="max-width:100%;height:auto;" /><figcaption>Figure 1. Histogram showing bimodal distribution of p-NF-κB signaling intensity in LPS-stimulated B cells</figcaption></figure><p>Further investigation into the drivers of this stochasticity revealed that the cell's baseline metabolic state was a significant predictor of activation capacity. Cells with higher levels of glycolytic enzymes (e.g., HK2) at the time of stimulation reached the signaling threshold more rapidly and exhibited higher peak phosphorylation levels (Table 2).</p><figure class="table-figure"><table><thead><tr><th>Immune Subset</th><th>Pathway</th><th>Activation Score (Mean)</th><th>Activation Heterogeneity (Fano Factor)</th></tr></thead><tbody><tr><td>Naive CD4+ T cells</td><td>TCR / CD3</td><td>0.42</td><td>1.85</td></tr><tr><td>Memory CD4+ T cells</td><td>TCR / CD3</td><td>0.78</td><td>1.12</td></tr><tr><td>B cells (CD19+)</td><td>TLR4 / LPS</td><td>0.55</td><td>2.34</td></tr><tr><td>Monocytes (CD14+)</td><td>TLR4 / LPS</td><td>0.89</td><td>0.95</td></tr></tbody></table><figcaption>Table 2. Pathway activation scores and heterogeneity across distinct immune cell subsets.</figcaption></figure><h4>Metabolic Reprogramming and Effector Function</h4><p>The integration of SCCP data with metabolic pathway analysis [9, 29] highlighted a distinct metabolic gradient within the activated cell population. We identified a sub-population of 'hyper-glycolytic' cells that was characterized by the overexpression of BHLHE40 and high levels of lactate dehydrogenase (LDH). These cells were the primary producers of pro-inflammatory cytokines, suggesting that metabolic commitment is a prerequisite for full effector function. Table 3 details the correlation between metabolic markers and functional outputs.</p><figure class="table-figure"><table><thead><tr><th>Metabolic Marker</th><th>Cytokine Output (IFN-γ) Correlation</th><th>Proliferative Index Correlation</th><th>Pathological Association</th></tr></thead><tbody><tr><td>HK2 (Hexokinase 2)</td><td>0.71</td><td>0.65</td><td>High in Autoimmunity [21]</td></tr><tr><td>LDHA (Lactate Dehydrogenase A)</td><td>0.68</td><td>0.58</td><td>High in NSCLC TME [1]</td></tr><tr><td>CPT1A (Fatty Acid Oxidation)</td><td>-0.34</td><td>-0.42</td><td>High in Quiescent Cells [9]</td></tr></tbody></table><figcaption>Table 3. Correlation of metabolic markers with functional and pathological states.</figcaption></figure><h4>Proteomic Signatures of Immune Aging</h4><p>In samples from elderly donors, we observed a shift in the proteomic landscape toward a senescent-associated secretory phenotype (SASP). Single-cell proteomic profiling identified a subset of 'inflammaging' T cells characterized by reduced TCR signaling responsiveness but elevated basal levels of p-p38 and p-NF-κB [8, 24]. This 'pre-activated' state appears to contribute to the chronic low-grade inflammation observed in aging populations.</p><figure class="article-figure"><img src="https://smnxsewcdnayrztrrghn.supabase.co/storage/v1/object/public/journal-assets/scholarly/single-cell-proteomics-unveiling-heterogeneity-in-immune-cell-activation-pathways-22h26/figure-2-1778839096450.png" alt="t-SNE visualization of proteomic clusters in young vs. elderly immune populations" loading="lazy" style="max-width:100%;height:auto;" /><figcaption>Figure 2. t-SNE visualization of proteomic clusters in young vs. elderly immune populations</figcaption></figure>
<h2>Discussion</h2>
<h4>The Functional Significance of Proteomic Heterogeneity</h4><p>Our findings demonstrate that proteomic heterogeneity is a fundamental characteristic of immune cell activation. The low correlation between mRNA and protein levels for signaling molecules (Table 1) underscores the limitations of relying solely on transcriptomic data to infer cellular function [11]. The stochasticity observed in NF-κB and MAPK activation suggests that immune responses are governed by probabilistic rather than deterministic rules at the single-cell level. This variability may serve as a 'bet-hedging' strategy, ensuring that a population of cells can respond to a wide range of stimulus intensities and durations [4, 6].</p><h4>Metabolic Priming as a Gatekeeper of Activation</h4><p>The correlation between baseline glycolytic enzyme levels and signaling magnitude (Table 3) suggests that metabolic state acts as a gatekeeper for immune activation. This supports the 'metabolic priming' hypothesis, where the availability of metabolic precursors and the activity of key enzymes like HK2 determine whether a cell can successfully transition from quiescence to an effector state [9, 29]. In disease contexts such as NSCLC and CRC, the tumor microenvironment may exploit this dependency by depleting essential nutrients, thereby dampening the activation potential of infiltrating immune cells [1, 3, 22].</p><h4>Clinical Implications and Precision Medicine</h4><p>The identification of specific proteomic sub-populations, such as the 'hyper-glycolytic' effector cells or the 'inflammaging' senescent cells, has significant clinical implications. These populations represent potential targets for precision therapies. For instance, metabolic inhibitors could be used to selectively dampen overactive pro-inflammatory subsets in autoimmune diseases [21, 26], while rejuvenating the metabolic capacity of T cells could enhance the efficacy of cancer immunotherapies [27, 29]. The use of SCNP and SCCP in clinical trials could provide real-time monitoring of treatment response at the single-cell level, allowing for more personalized therapeutic adjustments [12, 19].</p><h4>Technical Limitations and Future Directions</h4><p>Despite the insights provided by single-cell proteomics, several challenges remain. The depth of the proteome captured per cell (approx. 1,500 proteins in this study) is still significantly lower than the coverage achieved by scRNA-seq. Improvements in MS sensitivity and microfluidic sample preparation are needed to reach the depth required for comprehensive pathway mapping [11]. Furthermore, integrating spatiotemporal information with proteomic data remains a complex task [30]. Future research should focus on multi-omic integration, combining proteomics with epigenomics and spatial imaging to fully resolve the regulatory logic of immune cell heterogeneity [17, 25].</p>
<h2>Conclusion</h2>
<p>Single-cell proteomics has unveiled a hidden layer of complexity in immune cell activation pathways, revealing that functional heterogeneity is driven by protein-level stochasticity and metabolic commitment. Our study demonstrates that transcriptomic profiles often fail to capture the immediate signaling events and metabolic shifts that define an immune cell's response to stimulus. By identifying the proteomic signatures of effector gradients, metabolic priming, and immune aging, we provide a deeper understanding of how cellular diversity contributes to both health and disease. As technologies like SCCP and SCNP continue to mature, they will become indispensable tools in the transition toward truly precision immunology, enabling the targeted modulation of specific cellular states to improve patient outcomes across a spectrum of immunological and oncological disorders.</p>
<h2>References</h2>
<ol class="references">
<li>Unknown. Dissecting heterogeneity and immune cell populations in non-small cell lung cancer by single-cell RNA sequencing. International Journal of Frontiers in Medicine. 2023;5(10). https://doi.org/10.25236/ijfm.2023.051013</li>
<li>Végvári, Á., Rodriguez, J. E., Zubarev, R.. Single Cell Chemical Proteomics (SCCP) Interrogates the Timing and Heterogeneity of Cancer Cell Commitment to Death. SSRN Electronic Journal. 2021. https://doi.org/10.2139/ssrn.3956651</li>
<li>Zhu, J., Lian, J., Lu, H.. Unveiling Cellular Heterogeneity in Colorectal Cancer Through Single-Cell Sequencing. International Journal of Molecular Medical Science. 2024. https://doi.org/10.5376/ijmms.2024.14.0026</li>
<li>Papalexi, E., Satija, R.. Single-cell RNA sequencing to explore immune cell heterogeneity. Nature Reviews Immunology. 2017;18(1), 35-45. https://doi.org/10.1038/nri.2017.76</li>
<li>Pavillon, N., Smith, N. I.. Immune cell type, cell activation, and single cell heterogeneity revealed by label-free optical methods. Scientific Reports. 2019;9(1). https://doi.org/10.1038/s41598-019-53428-3</li>
<li>Boutros, M., Agaisse, H., Perrimon, N.. Sequential Activation of Signaling Pathways during Innate Immune Responses in Drosophila. Developmental Cell. 2002;3(5), 711-722. https://doi.org/10.1016/s1534-5807(02)00325-8</li>
<li>Qin, L.. Studying Tumor Initiating Cell (TIC) Heterogeneity and Treatment Resistance by Single-Cell Microfluidics Technology. Single Cell Biology. 2012;01(03). https://doi.org/10.4172/2168-9431.1000e112</li>
<li>Chan, C. L., Sugimura, R.. Unveiling the immune system aging in single-cell resolution. Journal of Leukocyte Biology. 2023;115(1), 16-18. https://doi.org/10.1093/jleuko/qiad136</li>
<li>Pearce, E., Pearce, E.. Metabolic Pathways in Immune Cell Activation and Quiescence. Immunity. 2013;38(4), 633-643. https://doi.org/10.1016/j.immuni.2013.04.005</li>
<li>Jain, S., Wang, E., Babikir, H., Shamardani, K., Diaz, A., Aghi, M. K.. EPCO-28. SINGLE-CELL RNA SEQUENCING IDENTIFIES INTRATUMORAL HETEROGENEITY AND ACTIVATION OF PRO-INVASIVE PATHWAYS IN NON-FUNCTIONING PITUITARY ADENOMAS. Neuro-Oncology. 2021;23(Supplement_6), vi8-vi8. https://doi.org/10.1093/neuonc/noab196.027</li>
<li>Yang, L., George, J., Wang, J.. Deep Profiling of Cellular Heterogeneity by Emerging Single‐Cell Proteomic Technologies. PROTEOMICS. 2019;20(13). https://doi.org/10.1002/pmic.201900226</li>
<li>Longo, D., Conroy, A., Louie, B., Wang, E., Pos, Z., Marincola, F.. Single cell network profiling (SCNP) assay allows for the identification of heterogeneity in the activation of multiple signaling pathways in distinct human B cell subsets (174.12). The Journal of Immunology. 2012;188(1_Supplement), 174.12-174.12. https://doi.org/10.4049/jimmunol.188.supp.174.12</li>
<li>Zhang, R., Liu, X., Ma, Y., Cheng, L., Ren, Y., Li, R.. Identification of Cell-Cell Communications by Single-Cell RNA Sequencing in End Stage Renal Disease Provides New Insights into Immune Cell Heterogeneity. Journal of Inflammation Research. 2023;Volume 16, 4977-5000. https://doi.org/10.2147/jir.s424911</li>
<li>Unknown. Innate Immune DNA-Sensing Pathways. Cell. 2013;152(3), 375-377. https://doi.org/10.1016/j.cell.2013.01.026</li>
<li>Ma, J., Zhang, L., Zhan, H., Mo, Y., Ren, Z., Shao, A.. Single-Cell Transcriptomic Analysis of Endometriosis Provides Insights Into Fibroblast Fates and Immune Cell Heterogeneity. SSRN Electronic Journal. 2021. https://doi.org/10.2139/ssrn.3780107</li>
<li>Unknown. Single-Cell Western Blotting to Determine Cell Heterogeneity. Journal of Visualized Experiments. 2023. https://doi.org/10.3791/21368</li>
<li>Do, T., Perrie, J., Pellegrini, M., Gudjonsson, J., Ma, F., Modlin, R.. 373 Identification of immune cell pathways in acne vs. rosacea by single-cell RNA sequencing and single-cell spatial imaging. Journal of Investigative Dermatology. 2023;143(5), S64. https://doi.org/10.1016/j.jid.2023.03.378</li>
<li>Unknown. Single-Cell Western Blotting to Determine Cell Heterogeneity. Journal of Visualized Experiments. 2023. https://doi.org/10.3791/21388</li>
<li>Chuan, C. H.. Proteomics in Clinical Studies and its Applications. Single Cell Biology. 2012;01(02). https://doi.org/10.4172/2168-9431.1000e108</li>
<li>Tong, Z., Mang, G., Wang, D., Cui, J., Yang, Q., Zhang, M.. Single-Cell RNA Sequencing Maps Immune Cell Heterogeneity in Mice with Allogeneic Cardiac Transplantation. Cardiovascular Innovations and Applications. 2023;8(1). https://doi.org/10.15212/cvia.2023.0023</li>
<li>Luo, S., Wang, L., Xiao, Y., Cao, C., Liu, Q., Zhou, Y.. Single-Cell RNA-Sequencing Integration Analysis Revealed Immune Cell Heterogeneity in Five Human Autoimmune Diseases. BIO Integration. 2023;4(4). https://doi.org/10.15212/bioi-2023-0012</li>
<li>Guinney, J., Dienstmann, R., Wang, X., Reyniès, A. d., Schlicker, A., Soneson, C.. The consensus molecular subtypes of colorectal cancer. Nature Medicine. 2015;21(11), 1350-1356. https://doi.org/10.1038/nm.3967</li>
<li>Ge, X., Son, E., Yao, R.. iDEP: an integrated web application for differential expression and pathway analysis of RNA-Seq data. BMC Bioinformatics. 2018;19(1), 534-534. https://doi.org/10.1186/s12859-018-2486-6</li>
<li>Saul, D., Kosinsky, R. L., Atkinson, E. J., Doolittle, M. L., Zhang, X., LeBrasseur, N. K.. A new gene set identifies senescent cells and predicts senescence-associated pathways across tissues. Nature Communications. 2022;13(1), 4827-4827. https://doi.org/10.1038/s41467-022-32552-1</li>
<li>Liu, Y., Sethi, N. S., Hinoue, T., Schneider, B., Cherniack, A. D., Sánchez-Vega, F.. Comparative Molecular Analysis of Gastrointestinal Adenocarcinomas. Cancer Cell. 2018;33(4), 721-735.e8. https://doi.org/10.1016/j.ccell.2018.03.010</li>
<li>Zhang, P., Zhang, D., Zhou, W., Wang, L., Wang, B., Zhang, T.. Network pharmacology: towards the artificial intelligence-based precision traditional Chinese medicine. Briefings in Bioinformatics. 2023;25(1). https://doi.org/10.1093/bib/bbad518</li>
<li>Zhang, Q., Lou, Y., Yang, J., Wang, J., Feng, J., Zhao, Y.. Integrated multiomic analysis reveals comprehensive tumour heterogeneity and novel immunophenotypic classification in hepatocellular carcinomas. Gut. 2019;68(11), 2019-2031. https://doi.org/10.1136/gutjnl-2019-318912</li>
<li>Cano-Gamez, K., Soskic, B., Roumeliotis, T. I., So, E. C., Smyth, D. J., Baldrighi, M.. Single-cell transcriptomics identifies an effectorness gradient shaping the response of CD4+ T cells to cytokines. Nature Communications. 2020;11(1), 1801-1801. https://doi.org/10.1038/s41467-020-15543-y</li>
<li>Wang, L., Liu, Y., Dai, Y., Tang, X., Tong, Y., Wang, C.. Single-cell RNA-seq analysis reveals BHLHE40-driven pro-tumour neutrophils with hyperactivated glycolysis in pancreatic tumour microenvironment. Gut. 2022;72(5), 958-971. https://doi.org/10.1136/gutjnl-2021-326070</li>
<li>Pham, D., Tan, X., Balderson, B., Xu, J., Grice, L. F., Yoon, S.. Robust mapping of spatiotemporal trajectories and cell–cell interactions in healthy and diseased tissues. Nature Communications. 2023;14(1), 7739-7739. https://doi.org/10.1038/s41467-023-43120-6</li>
</ol>
</article>