Upload your resume. PhdFit's six research agents compare your background with faculty profiles, recent publications, lab focus, and outreach opportunities, then rank professors with evidence you can review.
Resume-aware matchingEvidence from recent researchOutreach drafts included
Advisor Match Brief Resume uploaded
Resume signals5 found
NLPHCIFairnessGraph MLPyTorch
Six-agent analysisRunning
Axel · Candidate AnalystResume parsed
Nova · Professor ResearcherRecent papers
Echo · Match ExplainerEvidence linked
Atlas · Strategy AdvisorFit ranked
Lyra · Outreach WriterDraft queued
Juno · Meeting Prep CompilerNext step ready
SC
Dr. Sarah Chen
Stanford · NLP & interpretability
91
fit
Fairness + NLP projects overlap with her recent work.
Strong methods fit: model evaluation and interpretability.
Clear outreach angle for current trustworthy AI research.
Outreach angle
Ask how her lab is extending interpretability methods into fairness audits for real-world AI systems.
Xiaoping Bao
· William K. Luckow Associate Professor of Chemical EngineeringVerified
The Bao research group applies chemical and cellular engineering approaches to differentiate and biomanufacturing human stem cell therapeutics targeting various diseases such as cardiovascular diseases and cancer. Our group is part of the Davidson School of Chemical Engineering.
Pancreatic ductal organoids (PDOs) generated from human induced pluripotent stem cells (iPSCs) can be used to model pancreatic diseases and to conduct drug screening/testing. However, current protocols for generating PDOs rely heavily on tumor-derived Matrigel, which has been shown to upregulate oncogenes. Furthermore, Matrigel has undefined composition and weak mechanical properties that hamper mechanistic studies of cell-material interactions. In this study, we explore photo-clickable decellularized small intestine submucosa-norbornene (dSIS-NB) hydrogels as a Matrigel replacement for generating human iPSC-derived PDOs. To achieve this, pancreatic progenitors (PP) were first differentiated in conventional two-dimensional (2D) culture, aggregated into spheroids, then encapsulated and differentiated within dSIS-NB hydrogels with tunable stiffness. The differentiated organoids were analyzed by morphology, expression of key pancreatic ductal markers, and single-cell RNA sequencing (scRNA-seq). Post-differentiation, PDOs generated in stiffer photo-clickable dSIS-NB hydrogels (shear moduli ∼2.5 kPa) maintained ductal epithelial phenotype and exhibited pronounced forskolin-induced swelling. In contrast, differentiation of PP spheroids in softer dSIS-NB gels (shear moduli ∼0.9 kPa) and Matrigel resulted in a persistent mesenchymal phenotype and failed to generate functional PDOs. Finally, scRNA-seq results revealed that stiffer dSIS-NB hydrogels strongly biased ductal cell differentiation, yielding greater than 97% ductal progeny.
Single-cell RNA sequencing (scRNA-seq) has transformed biomedical research by enabling transcriptomic analysis at single-cell resolution. Yet, existing computational approaches remain primarily data-driven and lack the ability to integrate research context, limiting their interpretability and impact on hypothesis generation or experimental planning. We present scChat, a large language model (LLM)-powered co-pilot for contextualized scRNA-seq analysis. Unlike conventional pipelines restricted to tasks such as cell type annotation or enrichment analysis, scChat has an interactive, reasoning-based framework. It combines quantitative algorithms with retrieval-augmented generation and a multi-agent architecture to support hypothesis validation, mechanistic interpretation, and next-step experimental design. Through showcase and benchmarking studies, we demonstrate that scChat not only achieves high accuracy in cell type annotation but also provides biologically grounded explanations and contextual insights.
Abstract Pancreatic ductal adenocarcinoma (PDAC) creates complex tumor microenvironment (TME) hallmarked with a desmoplastic stroma that facilitates tumor growth/invasion, chemoresistance, and immunosuppression. It urgently needs the identification and evaluation of stromal components that can be targeted to reprogram the stroma to improve drug delivery and efficacy without making tumors more aggressive. Thus, we hypothesize that the coagulation system in the PDAC TME can be targeted to reprogram PDAC stroma to alleviate chemoresistance and drug delivery barriers. Specifically, the thrombin/protease-activated receptor 1 (PAR1) signaling axis can be targeted to suppress growth/invasion of pancreatic cancer cells (PCCs) and cancer associated fibroblast (CAF)-derived fibrosis. Our underlying rationale is based upon a leaky tumor vasculature in PDAC resulting in the release of circulating coagulation factors and subsequent activation of the coagulation system in the TME. Tissue factor expressed by PCCs initiates the conversion of prothrombin to the active serine protease thrombin, which then activates PAR1, whose signaling is thought to promote PCC growth/invasion and CAF-mediated fibrosis. We developed and employed novel microphysiological systems (MPS) of PDAC tumor-stroma, which were designed to reconstitute extravascular coagulation in the PDAC TME to specifically investigate the role of thrombin-PAR1 signaling events on PCC growth and CAF-mediated fibrosis. Our MPS was a microfluidic platform where PCC and CAF were co-cultured in the 3D extracellular matrix perfused with/without thrombin. In addition, PAR1 expression in murine and human PCCs and CAFs was genetically modified or pharmacologically inhibited. Our MPS enabled systematic and translational analyses on the therapeutic potential of blocking PAR1 signaling in PCCs, CAFs, or both. In murine MPS, genetic deletion of PAR1 drastically decreased thrombin-mediated PCC and CAF growth compared to that of MPS with wildtype cells. Human MPS with varying levels of PAR1 also suggest thrombin stimulates PCC-CAF crosstalk, including CAF growth, elevated expression of a-SMA and secreted collagen levels. Furthermore, pharmacological inhibition of PAR1 by vorapaxar decreases both PCC and CAFs in all human PCC/CAF pairs studied. Finally, we confirm the findings from our MPS using PDAC tumor-stroma xenograft models with both human PCC and CAF. A significant reduction in tumor size is observed with vorapaxar treatment, which attributes primarily to the reduction of CAFs. In summary, we validate and translate the therapeutic potential of thrombin-PAR1 signaling in reprogramming PDAC stroma using novel MPS of PDAC tumor-stroma model. Our study also demonstrates MPS as a promising system for target identification, validation, and streamlining preclinical studies for drug discovery. Citation Format: Sae Rome. Choi, Hye-ran Moon, Natalia Ospina Muñoz, Yun Chang, Xiaoping Bao, Bennett D. Elzey, Meliss L. Fishel, Matthew J. Flick, Bumsoo Han. Validation and translation of therapeutic potential of thrombin-PAR1 signaling in suppressing fibrosis using microphysiological PDAC tumor models [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Pancreatic Cancer Research—Emerging Science Driving Transformative Solutions; Boston, MA; 2025 Sep 28-Oct 1; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2025;85(18_Suppl_3):Abstract nr B037.
Polymorphonuclear neutrophils (PMNs), the most abundant leukocytes circulating in human blood, are pivotal players in the innate immune system. In recent years, PMNs have gained increasing recognition for their significant involvement in the pathogenesis of a wide array of human diseases, including sepsis, pulmonary conditions, autoimmune disorders, and various cancers. Due to their terminally differentiated state, PMNs possess a short lifespan and exhibit limited proliferative potential, which makes continuous replenishment from the bone marrow essential for maintaining immune homeostasis. This demand underscores the need for efficient, reliable, and robust methods of PMN production. In this study, we evaluated three forward programming protocols and one directed differentiation protocol aimed at generating PMNs from human pluripotent stem cells (hPSCs). We analyzed not only their differentiation efficiency but also the transcriptomic profiles and functional capabilities of the resulting PMNs. Our findings revealed that both the forward programming method and the directed differentiation approach can successfully generate functional PMNs. Furthermore, by fine-tuning the culture media at various stages during forward programming, we identified an optimal protocol that significantly enhances hematopoietic differentiation potential and promotes the functional maturity of the neutrophils.
bioRxiv (Cold Spring Harbor Laboratory) · 2025-04-09 · 4 citations
preprintOpen access
ABSTRACT Human pluripotent stem cell (hPSC)-derived cardiac therapies hold great promise for heart regeneration but face major translational barriers due to allogeneic immune rejection. Here, we engineered hypoimmunogenic hPSCs using a two-step CRISPR-Cas9 strategy: (1) B2M knockout, eliminating HLA class I surface expression, and (2) knock-in of HLA-E or HLA-G trimer constructs in the AAVS1 safe harbor locus to confer robust immune evasion. Hypoimmunogenic hPSCs maintained pluripotency, efficiently differentiated into cardiac cell types that resisted both T and NK cell-mediated cytotoxicity in vitro , and self-assembled into engineered cardiac organoids. Comprehensive analyses of the hypoimmunogenic cells and organoids revealed preservation of transcriptomic, structural, and functional properties with minimal off-target effects from gene editing. In vivo , hypoimmunogenic cardiac organoids restored contractile function in infarcted rat hearts and demonstrated superior graft retention and immune evasion in humanized mice compared to wild-type counterparts. These findings establish the therapeutic potential of hypoimmunogenic hPSC-CMs in the cardiac organoid platform, laying the foundation for off-the-shelf cardiac cell therapies to treat cardiovascular disease, the leading cause of death worldwide.
Cancer immunotherapy, specifically Chimeric Antigen Receptor (CAR)-T cell therapy, represents a significant breakthrough in treating cancers. Despite its success in hematological cancers, CAR-T exhibits limited efficacy in solid tumors, which account for more than 90% of all cancers. Solid tumors commonly present unique challenges, including antigen heterogeneity and complex tumor microenvironment (TME). To address these, efforts are being made through improvements in CAR design and the development of advanced validation platforms. While efficacy is limited, some solid tumor types, such as neuroblastoma and gastrointestinal cancers, have shown responsiveness to CAR-T therapy in recent clinical trials. In this review, it is first examined both experimental and computational strategies, such as protein engineering coupled with machine learning, developed to enhance T cell specificity. The challenges and methods associated with T cell delivery and in vivo reprogramming in solid tumors is discussed. It is also explored the advancements in engineered organoid systems, which are emerging as high-fidelity in vitro models that closely mimic the complex human TME and serve as a validation platform for CAR discovery. Collectively, these innovative engineering strategies offer the potential to revolutionize the next generation of CAR-T therapy, ultimately paving the way for more effective treatments in solid tumors.
Adoptive immune cell-based therapies have shown promise in cancer treatment, yet their efficacy against solid tumors is often limited by the immunosuppressive tumor microenvironment (TME). To overcome these barriers, we design an innovative immune cell cocktail as a combinatorial biomaterial platform, which harnessing the complementary functions of neutrophils and natural killer (NK) cells derived from human pluripotent stem cells (hPSCs). Using CRISPR/Cas9, we introduce an anti-fluorescein isothiocyanate (FITC) chimeric antigen receptor (CAR) construct into the AAVS1 safe harbor locus of hPSCs, allowing for the differentiation of CAR-modified neutrophils and NK cells. These CAR neutrophils exhibit robust anti-tumor activity, forming immune synapses with tumor cells tagged via a bispecific adapter (FITC-folate), even in hypoxic TMEs, while CAR NK cells demonstrate antigen-specific cytotoxicity. Together, the cocktail biomaterial composed of CAR neutrophils and CAR NK cells creates a synergistic anti-tumor effect: having neutrophils enhance TME modulation, and NK cells provide targeted cytotoxicity. This biomaterial offers a scalable and off-the-shelf solution for producing CAR neutrophils and CAR NK cells, potentially reducing needs for high-dose exogenous cytokines and minimizing immune-related toxicities. Our findings suggest that hPSC-derived CAR neutrophils and CAR NK cells may form an effective immuno-cocktail biomaterial, offering a feasible strategy for advancing solid tumor immunotherapy through cellular synergy and TME adaptation. • CAR neutrophils and NK cells from hPSC offer scalable and off-the-shelf adoptive immune cell therapy solutions. • CAR neutrophil-NK cell cocktails boost synergistic antitumor effects, offering a promising immune cell therapy strategy. • FITC-folate bridged CAR immune cocktails boost targeting and function, offering a new strategy for solid tumor therapy.