NCI Division of Cancer Biology Junior Investigators Meeting (Abstract)

Filter submissions
14 submissions
# Created User IP address First Name Middle Initial Last Name Degree(s) Position/Title/Career Status Organization Organization Address Email Abstract Category Abstract Keywords Abstract Title Abstract Operations
14 Tue, 09/08/2026 - 17:28 Anonymous 10.208.24.147 Ali Mohammad M.Sc. PhD Candidate University of Minnesota Minneapolis moha1722@umn.edu Consider my abstract for a Methodology/Technology presentation T cell activation, Cell volume regulation, Amino acid transport, Protein synthesis, Cell density Mechanistic modeling of water uptake and biomass accumulation during CD8⁺ T cell growth following activation Abstract:
Experimental measurements show that resting CD8⁺ T cells increase their volume approximately fourfold within 48 hours of activation while simultaneously accumulating biomass and increasing their fractional water content, resulting in a decrease in cell density. To investigate how activated T cells coordinate water uptake and biomass accumulation, we developed a mechanistic ODE-based Cell Growth Simulator that couples ion transport, osmotically driven water flux, amino acid transport, and protein biosynthesis during the first 48 hours following activation. The model tracks ten state variables representing cellular water volume, intracellular inorganic ion concentrations, free amino acid concentrations, and macromolecular mass, and is calibrated using high-resolution single-cell measurements of volume and density obtained with the suspended microchannel resonator (fxSMR; Wu et al., Nature Biomedical Engineering, 2025) together with intracellular free amino acid concentrations from metabolomic analysis (O’Keeffe et al., bioRxiv, 2025).
To determine whether the amino acid transport is uniquely required or whether alternative mechanisms could explain the observed growth, we performed a structural non-identifiability analysis across five candidate ion-transport models. Although multiple mechanisms could be tuned to reproduce the measured volume trajectory, all failed to simultaneously reproduce density, missing the experimental buoyant mass target by 40–65% despite correctly matching volume. Mechanistically, ion-driven growth increases water content without proportionally increasing dry mass (primarily protein), yielding cells of the correct size but incorrect composition. In contrast, a model further incorporating amino-acid-transport and protein biosynthesis satisfied all experimental observables. These results establish cell density as critical constraint for discriminating among competing growth mechanisms and provide a quantitative framework for linking T cell metabolic activation to its biophysical growth phenotype. Our integrated model contributes to the identification of metabolically and biophysically motivated targets for accelerating T cell expansion for therapeutic benefit.
13 Tue, 09/08/2026 - 13:53 Anonymous 10.208.24.147 Sai Ma Ph.D. Assistant Professor Icahn School of Medicine at Mount Sinai New York sai.ma2@mssm.edu Consider my abstract for a Methodology/Technology presentation Single-cell, multi-comics, gene regulation Scalable Single-Cell Multimodal Genomics for Mapping Regulatory State Transitions Understanding how regulatory programs change during cell-state transitions requires technologies that can measure multiple layers of genome regulation in the same cell and at sufficient scale to capture rare and transient states. We are developing a suite of scalable single-cell multimodal genomic technologies designed to connect epigenetic regulation, genome organization, and transcription within individual cells. These methods enable ultra-high-throughput profiling of 100,000 to 1 million cells per assay, providing the statistical power and cellular resolution needed to resolve rare populations, reconstruct continuous state transitions, and systematically interrogate regulatory heterogeneity.
A major focus is ME-seq, a combinatorial-indexing platform that jointly profiles DNA methylation, chromatin accessibility, and gene expression from the same cell. By integrating enzymatic methylation profiling with highly scalable indexing, ME-seq enables trimodal measurements across hundreds of thousands of cells while preserving regulatory information. We have applied ME-seq across developmental, hematopoietic, aging, and disease systems, where same-cell measurements allow us to distinguish regulatory changes that precede transcriptional state transitions from those that accompany or follow them.
We are also extending this framework to incorporate three-dimensional genome organization through technologies that jointly measure chromatin contacts and transcription at single-cell resolution. These approaches are being optimized for increased molecular recovery, high-throughput processing, and compatibility with targeted enrichment strategies, enabling scalable interrogation of genome architecture in heterogeneous primary cell populations.
Together, these technology-development efforts aim to move single-cell genomics beyond parallel molecular atlases toward direct measurement of regulatory coupling within the same cell. By combining ultra-high throughput with increasingly comprehensive multimodal measurements, these platforms enable analysis of rare cellular states, regulatory trajectories, cell-to-cell heterogeneity, and coordinated molecular changes that would be difficult to resolve using conventional-scale single-cell assays. These capabilities provide new opportunities to identify early regulatory events preceding phenotypic change and dissect mechanisms of cellular transitions in development and cancer.
12 Tue, 09/08/2026 - 13:24 Anonymous 10.208.24.147 Ruohan Wang Ph.D. Stanford University Stanford ruohwang@stanford.edu Use my abstract for team formation only (do not consider it for a presentation) Cancer Immunotherapy; Tumor microenvironment; Deep learning Deep-learning–driven tumor microenvironment profiling improves immunotherapy response prediction Immunotherapy benefits only a fraction of cancer patients, and we still lack reliable ways to predict who will respond. A key reason is that response depends not on any single biomarker but on the collective behavior of diverse cell populations within the tumor microenvironment (TME), communicating through structured signaling networks that existing methods do not capture.
We developed EcoNet, a framework that first profiles the TME as co-occurring cell-state ecosystems (ecotypes), then reconstructs the intercellular and intracellular signaling networks connecting them, and feeds these networks into a graph attention model to predict immunotherapy outcome from bulk RNA-seq. Pretrained on 2,532 pan-cancer immunotherapy samples, EcoNet outperformed existing prediction tools by 7.5 to 19.9% and marker-based signatures by 4.4 to 21.2% in AUC. Its high-attention genes were independently validated by a CRISPR screen for T-cell-mediated tumor killing. Applied to clear cell renal cell carcinoma with disease-specific fine-tuning, EcoNet stratified an independent cohort by progression-free survival (P=0.024) and identified an interferon-responsive program linking antigen presentation, immune infiltration, and checkpoint regulation. We further validated these findings using CosMx single-cell spatial transcriptomics on a 46-sample clinical cohort (~450,000 cells), confirming that the therapy-beneficial TME configuration localizes to immune-enriched tissue niches.
EcoNet provides a generalizable, biologically grounded approach to predicting immunotherapy response and is readily extensible to other cancers and therapeutic settings.
11 Tue, 09/08/2026 - 12:16 Anonymous 10.208.28.41 Elham Seyyedi Zadeh Ph.D. Graduate student Auburn University Auburn ezs0122@auburn.edu Consider my abstract for a Methodology/Technology presentation Lung cancer, Tumor microenvironment; Cancer-associated fibroblasts; PEG-fibrinogen hydrogels; High throughput screening A Tunable 3D Hydrogel Platform for Modeling Stromal-Driven Tumor Behavior My research is built on a simple premise: to understand a tumor, you have to study more than its cancer cells. As a PhD candidate in Chemical Engineering at Auburn University, working under Dr. Elizabeth Lipke, I build 3D tissue-engineered models of the tumor microenvironment. My focus is on non-small cell lung cancer and colorectal cancer, and how stromal cells, particularly cancer-associated fibroblasts, shape tumor mechanics, structure, and drug response.

My core technology is a poly(ethylene glycol)–fibrinogen (PEG-Fb) hydrogel platform. It lets me build 3D co-culture tissues with precisely tunable cancer-to-stromal cell ratios, from fibroblast-poor to fibroblast-rich. I apply this platform across established cell lines, patient-derived xenografts, and patient-derived organoids, adapting each protocol to the model at hand. Around this platform, I have built a characterization toolkit: live/dead viability assays with image-based quantification; phase-contrast and fluorescence microscopy; mechanical testing by parallel-plate compression to quantify tissue stiffness; flow cytometry and microfluidic fabrication of uniform tumor microspheres for high throughput drug screening.

What distinguishes my work is its collaborative core. Our computational collaborators at the University of Minnesota use single-cell and spatially resolved transcriptomic data to predict fibroblast-associated drug response, then test these predictions in 2D coculture. My role is to validate these predictions in 3D, which better recapitulates the mechanical and structural features of real tumors than flat culture.
10 Tue, 09/08/2026 - 09:06 Anonymous 10.208.28.41 Anna Michmerhuizen Ph.D. Postdoctoral Research Associate University of North Carolina at Chapel Hill Chapel Hill, NC anna_michmerhuizen@med.unc.edu Use my abstract for team formation only (do not consider it for a presentation) breast cancer, metastasis, metastatic tropism, computational modeling, precision oncology Identifying drivers of organ-specific breast cancer metastasis Metastatic breast cancer typically progresses to grow in critical organs and is the primary cause of breast cancer-related deaths. We and others have shown that organotrophic metastatic behavior is associated with the molecular subtype of the primary tumor. Specifically, the basal-like “intrinsic” subtype metastasizes preferentially to the lung and brain, HER2-enriched subtype to the liver, and luminal subtypes to the bone. Using data from 2486 human breast tumors where many have progressed to metastases (bone, brain, liver, lung, or multi-metastases [2+ sites]), we performed supervised machine learning to develop logistic regression models predictive of site-specific metastasis employing elastic net regularization for selection of interpretable features and model building. First, models predictive of distant metastasis were successfully built to classify patients based on whether their tumor will metastasize (AUC = 0.67). We next trained models predictive of site-specific metastasis, including a luminal bone metastasis model (AUC = 0.75), where selected features identify patients with bone metastases only (i.e. no visceral metastatic sites [brain, liver, lung]). Tumor cell intrinsic and microenvironment extrinsic features were selected in the final model. A model for multi-metastases was also developed (AUC = 0.68), and features associated with multi-site metastasis include hypoxia, clinical HER2 status, correlation to the HER2-enriched molecular subtype, and lymphovascular invasion. These overall modeling results suggest two distinct metastatic phenotypes (bone and viscera), which are supported by the performance of published site-specific metastasis signatures. Our models of site-specific metastasis may provide clinically useful biomarkers and nominate unique features for mechanistic studies.
9 Thu, 09/03/2026 - 12:19 Anonymous 10.208.24.244 Seongyeol Park M.D. Ph.D. Postdoc Stanford University Stanford, CA syparkmd@stanfrod.edu Consider my abstract for a Methodology/Technology presentation Decoding Cancer Evolution: Integrating Genomic, Transcriptomic, and Spatial Approaches to Intrinsic and Extrinsic Drivers I am a computational biologist with a strong interest in cancer evolution, particularly in how intrinsic and extrinsic factors shape the evolutionary trajectories of tumors. My research aims to understand how genomic instability, mutational processes, and the tumor microenvironment jointly influence cancer progression and clonal dynamics.

Over the past ten years, I have developed extensive expertise in DNA sequencing data analysis, including the identification and characterization of point mutations, structural variations, copy number alterations, and mutational signatures. This experience has given me a solid foundation for dissecting the genomic landscape of tumors and interpreting the evolutionary processes underlying cancer development.

In addition to genomic analysis, I have substantial experience in transcriptomic profiling, including bulk RNA sequencing and single cell RNA sequencing, which has allowed me to investigate gene expression heterogeneity both within tumors and across their surrounding cellular context.

More recently, my research has focused on cancer and tumor microenvironment (TME) interactions, with a particular emphasis on spatial transcriptomics. Using spatial transcriptome analysis, I aim to reveal how the spatial organization of tumors affects cell to cell communication, immune evasion, and clonal selection.

By integrating genomic, transcriptomic, and spatial data, I seek to build a comprehensive picture of how cancer evolves under selective pressures arising from both intrinsic genomic alterations and extrinsic factors within the tumor microenvironment. I look forward to discussing this integrative approach and exploring collaborative opportunities related to cancer evolution and TME research at this conference.
8 Tue, 09/01/2026 - 11:26 Anonymous 10.208.24.244 Daniel N Phipps MS PhD Candidate Kashatus Lab/University of Virginia Charlottesville, VA pux6qa@virginia.edu Use my abstract for team formation only (do not consider it for a presentation) Mitochondrial dynamics, colorectal cancer Drp1 promotes tumor cell fitness in colorectal cancer through mechanisms independent of mitochondrial fission. Colorectal cancer (CRC) ranks as the second leading cause of cancer-related deaths in the United States, and the top cause for cancer-related deaths in patients under 50. Approximately 40% of CRC patients carry an activating mutation in KRAS, a potent driver of oncogenic metabolism and cell replication. Hyperactive KRas signaling is known to promote dynamic structural changes in mitochondrial networks via ERK-mediated activation of the mitochondrial fission GTPase Drp1. Activated Drp1 has been identified as an important player in tumor cell proliferation and survival in RAS-driven disease. However, it remains unclear as to whether the pro-tumorigenic effects of Drp1 are directly attributable to changes in mitochondrial shape, or to presently unknown roles for Drp1 in tumor cell physiology. To address this question, our lab has developed two independent in vitro systems in which we express a fission-incompetent Drp1 in CRC cell lines. Surprisingly, while these cells are devoid of mitochondrial fission, the expression of fission-incompetent Drp1 significantly increases colony formation in comparison to Drp1-null cells. This exciting finding has led us to hypothesize that it is the localization of Drp1 to the outer mitochondrial membrane that induces pro-tumorigenic signaling and metabolism, rather than its canonical function in dividing mitochondria.
7 Mon, 08/31/2026 - 11:11 Anonymous 10.208.24.244 Russell Hawes B.S. Graduate Student University of Virginia Charlottesville rbh2equ@virginia.edu Use my abstract for team formation only (do not consider it for a presentation) A Generalized Additive Model of Location and Scale Isolates Cell-state Variation from Confounding Factors Differentiating meaningful biological variation from noise is a fundamental problem in single-cell transcriptomics. As an alternative for cell-state identification, previous studies employed stochastic 10-cell sequencing, a mini-bulk method that compares 10-cell samples to larger pooled samples to identify heterogeneous expression states while maintaining reproducibility. However, 10-cell sequencing alone cannot combine samples across batch variables, such as sex or individual patients, which reduces analytical sample sizes and limits biological insight. To improve this method controlling for batch effects and technical noise, we applied a generalized additive model of location and scale. This model treats observed variance as a linear combination of factor variances, such as those from technical error and batch effect, and isolates the residual, biologically meaningful variance. We applied this model to stochastic 10-cell sequencing data from a mouse model of gliomagenesis and to human estrogen receptor-positive breast cancer samples. In each setting, the generalized additive model of location and scale predicted heterogeneously expressed genes that previous analyses missed. We validated these model predictions using RNA fluorescence in situ hybridization in independent samples. The generalized additive model of location and scale increases biological information extracted from limited sample sizes. This model may be applicable to other mini-bulk methods seeking to isolate meaningful data from noise.
6 Wed, 08/26/2026 - 22:28 Anonymous 10.208.28.32 Florencia Picech Ph.D. Associate Research Scientist Columbia University New York fp2447@cumc.columbia.edu Use my abstract for team formation only (do not consider it for a presentation) A single-cell atlas identifies distinct programs of metastatic progression across prostate cancer subtypes Metastasis is the major cause of mortality in patients with advanced prostate cancer, yet the mechanisms underlying metastatic progression and treatment response remain poorly understood. Androgen receptor (AR) signaling is central to prostate tumorigenesis being the main therapeutic target. However, metastatic prostate cancer encompasses distinct biological subtypes that retain AR dependence or evolve toward AR independence, raising questions about how these divergent states influence metastatic progression and adaptation.
To investigate this heterogeneity, we generated genetically engineered mouse models (GEMMs) recapitulating alterations in key genes and pathways frequently dysregulated in human metastatic disease, including Pten, p53, Rb1, DNA damage response, and MAPK/RAS signaling. These models display diverse metastatic phenotypes, including bone metastasis, and both AR-high and AR-low disease in an immunocompetent background. Single-cell RNA sequencing across 181 samples from normal prostate, prostate tumors, and metastases from multiple anatomical sites generated an atlas of tumor, stromal, and immune cell populations.
The atlas revealed distinct tumor-intrinsic programs associated with AR status. AR-high tumors were enriched for epithelial-to-mesenchymal transition and osteogenic programs, consistent with osteomimicry, and exhibited a hypoxic, glycolytic profile. In contrast, AR-low tumors showed enrichment of neurodifferentiation and neurotransmitter signaling, consistent with neuromimicry, together with increased oxidative phosphorylation. These features were maintained in metastases, suggesting distinct mechanisms of adaptation. Furthermore, the tumor microenvironment was characterized by myeloid dominance across all metastatic tumors, with distinct neutrophil populations identified across subtypes.
Together, these findings demonstrate coordinated tumor-intrinsic and microenvironmental adaptations during prostate cancer metastasis establishing a platform to identify therapeutic vulnerabilities across disease subtypes.
5 Wed, 08/26/2026 - 12:09 Anonymous 10.208.24.244 Aakash Saha Ph.D. Postdoctoral Research Scientist Columbia University New York as7656@columbia.edu Consider my abstract for a Methodology/Technology presentation Protein–protein interactions, Structural bioinformatics, Parallel computing, Cancer signaling networks, Structure-based interaction prediction A User-Accessible Python-Based PrePPI Framework for Protein-Protein Interaction Prediction Protein–protein interactions (PPIs) underlie cellular processes, yet proteome-scale identification of functional interactions remains computationally challenging. All-to-all analyses may require evaluation of billions of protein and domain pairs, a burden amplified when considering interactions across species, including hosts and pathogens. PrePPI makes proteome-scale searches for binary interactions tractable by using evolutionarily conserved structural cues to prioritize pairs that can be modeled with known interaction templates.
PrePPI-SM evaluates interactions between structured protein regions using structural similarity, interface coverage, and conserved contacts within a naïve Bayesian framework. PrePPI-SLiM identifies interactions in which a structured domain recognizes a short peptide-like motif in a partner protein.
We rewrote the Perl-based PrePPI-SM pipeline as an end-to-end Python workflow and redesigned Skan, its structural-neighbor search, for scalable execution. Skan partitions a database of representative protein structures into balanced shards and splits large sequence clusters to prevent bottlenecks. A shared queue dynamically assigns these shards among 30 CPUs as workers become available, improving utilization across uneven workloads. The original workflow repeatedly searched directories and opened small interface files, creating input/output overhead. We consolidated these data into indexed SQLite databases so workers retrieve only required records, reducing file-system traffic and unnecessary data loading. Across 447 human proteins and domains, the redesigned structural-neighbor search achieved a median 13.4-fold end-to-end speedup without altering the structural scoring framework. Now comparable to Foldseek in runtime, Skan complements Foldseek’s efficient close-match search by capturing more diverse structural neighbors. This helps identify a binary complex with an appropriate interface for predictive modeling.
These approaches could contribute to collaborative cancer research by prioritizing interactions involving cancer-associated proteins, connecting poorly characterized proteins to signaling networks, and generating structural hypotheses for experimental testing. This workflow could complement experimental proteomics and functional studies, helping collaborators select candidate interactions and interpret their potential roles in cancer biology.
4 Thu, 08/13/2026 - 11:08 Anonymous 10.208.28.62 Jeffrey Hsu B.S. M.S. University of Virginia Charlottesville uem3ed@virginia.edu Use my abstract for team formation only (do not consider it for a presentation) HER2 locally modulates the hormone receptor-negative DCIS microenvironment Ductal carcinoma in situ (DCIS) breast premalignancies account for ≥20% of all new diagnoses yet they do not align with standard invasive breast cancer subtypes. In DCIS, amplification of the HER2 receptor tyrosine kinase is two times more common than in invasive breast cancer, suggesting distinct roles in premalignancy. Using highly multiplexed imaging and biocomputational statistics to profile HER2-associated DCIS microenvironments, we asked whether HER2 expression coincides with different cellular constituents around DCIS lesions that are hormone receptor (HR)–negative and immune infiltrated. HER2 protein expression was variable across and within patient samples and was associated with the largest change in stromal cellularity relative to other profiled DCIS cell states (KRT5, γH2AX, Ki67). Compared to DCIS lesions with low or absent HER2 (HER2-low), lesions with high HER2 expression (HER2-high) had significantly fewer fibroblasts, CD8+ T cells, and macrophages. Interestingly, co-variation analysis of cell-type pairs revealed increased associations between cancer-associated fibroblast (CAF)/myofibroblastic CAF (myCAF) densities and resting fibroblast/myofibroblast densities, suggesting tighter coupling among fibroblast states in HER2-high regions. Separately, we compared HER2-positive and HER2-negative HR–negative DCIS patient cases from the Human Tumor Atlas Network by differential expression analysis of secretome genes, identifying a depletion in the matrisome-associated gene, lysyl oxidase (LOX), in HER2-positive cases. In vivo studies suggested that periductal CAF and myCAF densities were depleted following HER2 overexpression, whereas LOX overexpression produced a reciprocal increase in these populations. Collectively, these results suggest that HER2 locally shifts fibroblast-state coordination and dilutes immune presence in the DCIS microenvironment through LOX.
3 Mon, 08/10/2026 - 17:11 Anonymous 10.208.28.62 Isha Bhorkar M.S. Graduate Student University of Michigan Ann Arbor, MI ibhorkar@umich.edu Consider my abstract for a Methodology/Technology presentation Mechanobiology, Shear stress, Microfluidics, Organoids Multiscale Approaches for Studying Fluid Shear Stress and Mechanoadaptation in Human Fallopian Tube Epithelium The fallopian tube (FT) epithelium is the tissue of origin for most high-grade serous ovarian carcinoma (HGSOC). FT epithelial cells are continuously exposed to fluid shear from follicular fluid release, peristaltic contractions, and ciliary beating, yet how this mechanical environment shapes epithelial behavior and early transformation risk remains poorly understood. My research addresses this gap by combining computational modeling of luminal fluid mechanics with experimental mechanosensing studies in patient-derived microfluidic models.
To characterize the physical environment, I use NanoCT imaging and a custom Python pipeline (skeletonization, perpendicular plane resampling, hydraulic diameter calculation) to reconstruct patient-specific FT luminal geometry and estimate physiological wall shear stress, capturing how mucosal folding shapes the flow environment. These geometric models also feed into computational simulations to generate patient-informed shear estimates.
To link this biomechanical context to cellular response, cells from patient FT organoids are exposed to calibrated, physiological shear stress in microfluidic devices, alongside immortalized comparator lines that provide a transformed counterpart to normal FT epithelium. Cellular responses are characterized through RNA sequencing, live calcium imaging with pharmacological modulation of mechanosensitive channels, and immunofluorescence-based quantification of cytoskeletal organization and epithelial morphology. A perfused, membrane-based device extends this workflow to resolve mechanosensing across ciliated and secretory cell populations.
This work has given me hands-on expertise in patient-derived organoid culture, microfluidic device design and fabrication, quantitative image analysis, and the integration of structural, computational, and transcriptomic data. This biophysical and quantitative perspective could contribute to collaborative projects that connect tissue mechanics, tissue architecture, and mechanosensing to early cancer risk, detection, or progression across tissue types.
2 Mon, 08/10/2026 - 14:07 Anonymous 10.208.28.62 Sydney Quijano B.S Graduate Student Purdue University West Lafayette Indiana squijano@purdue.edu Consider my abstract for a Methodology/Technology presentation Loss of SPINK5 mediates the outgrowth of disseminated breast cancer Following establishment and treatment of primary disease, systemically disseminated breast cancer cells can remain dormant for years only to evolve into therapy-resistant metastases. Specific environmental influences that contribute to the transition from dormant to metastatic disease remain poorly characterized. Herein, we utilized mouse models to evaluate the influence of alcohol consumption on emergence from pulmonary dormancy. Consistent with prior reports, we found that animals consuming alcohol demonstrated decreased adaptive immune presence and increased extracellular matrix deposition in their lungs compared to non-drinking counterparts. Seeding the D2.OR model of pulmonary dormancy cells followed by alcohol consumption caused outgrowth of macroscopic lesions. Isolation and culture of these tumors produced an independent subline (D2.OR-EtOH) capable of pulmonary outgrowth upon reinjection into immunocompetent, alcohol naïve animals. Gene expression analysis of D2.OR-EtOH indicated downregulation of the serine peptidase inhibitor Kazal type 5 (SPINK5) compared to parental cells. Genetic depletion of SPINK5 from parental D2.OR cells was sufficient to allow pulmonary outgrowth in vivo. Evaluation of the immune infiltrate into alcohol-induced pulmonary lesions demonstrated high numbers of neutrophils and the formation of neutrophil extracellular traps (NETs), structures associated with breakage of pulmonary dormancy. D2.OR-EtOH subline and SPINK5-depleted D2.OR cells more readily induced NET formation when cocultured with neutrophils. Overall, our studies demonstrate that alcohol consumption can break pulmonary dormancy by altering the interaction of tumor cells with innate immune cells. Furthermore, our work identifies SPINK5 as an important contributor to metastatic progression in breast cancer through its regulation of interacting neutrophils.
1 Mon, 08/10/2026 - 10:54 Anonymous 10.208.24.93 Heber L. Rocha Ph.D. in Computational Modeling Assistant Scientist Indiana University Bloomington hlimadar@iu.edu Consider my abstract for a Methodology/Technology presentation agent-based modeling, uncertainty quantification, PhysiCell, model calibration, Bayesian inference, tumor microenvironment, computational oncology, open-source software UQ-PhysiCell: Uncertainty Quantification and Calibration for Agent-Based Models of Cancer My research in computational and mathematical oncology centers on multiscale agent-based models (ABMs) of cancer, in which individual cells are autonomous agents governed by rules for proliferation, migration, death, and signaling, coupled to continuum descriptions of oxygen, nutrients, and cytokines. As one of the developers of PhysiCell, an open-source framework for physics-based multicellular simulation, I have built ABMs of tumor growth, immune-tumor interactions, and treatment response that reproduce experimentally observed behaviors. However, the high-dimensional parameter spaces, stochasticity, and computational cost of ABMs pose major challenges for calibration, uncertainty quantification (UQ), and systematic comparison of competing mechanistic hypotheses, capabilities that are essential for predictive, data-grounded modeling. The next stage of my research addresses this gap. I lead development of UQ-PhysiCell, an extensible open-source Python framework that enables uncertainty quantification, calibration, and model selection for PhysiCell models. UQ-PhysiCell manages simulation inputs and outputs (parameters, initial conditions, cell behavior rules) and orchestrates large simulation ensembles with multiple levels of parallelism, integrating directly with established Python libraries for sensitivity analysis, optimization, Bayesian inference, and surrogate modeling. By decoupling model execution from statistical analysis, it lowers the barrier to rigorous, reproducible uncertainty-aware analysis and moves ABMs beyond single best-fit simulations toward interpretable predictions with quantified confidence. In collaborative cancer research projects, I contribute end-to-end modeling support: formalizing mechanistic hypotheses, building and calibrating tumor microenvironment ABMs against imaging and spatial omics data, and quantifying uncertainty in model-based predictions to guide experimental design.