Siyu He

All Publications

• A foundation generative model for breast ultrasound image analysis. Nature biomedical engineering Yu, H., Li, Y., Zhang, N., Niu, Z., Gong, X., Luo, Y., Ye, H., He, S., Wu, Q., Qin, W., Zhou, M., Han, J., Tao, J., Zhao, Z., Dai, D., He, D., Wang, D., Tang, B., Huo, L., Zou, J., Zhu, Q., Wang, Y., Wang, L. 2026

  Abstract

  Foundation models have emerged as powerful tools for addressing various tasks in clinical settings. However, their potential development for breast ultrasound analysis remains untapped. Here we present BUSGen, the first foundation generative model designed for breast ultrasound image analysis. Pretrained on over 3.5 million breast ultrasound images, BUSGen has acquired extensive knowledge of breast structures, pathological features and clinical variations. With few-shot adaptation, BUSGen can generate repositories of realistic and informative task-specific data, facilitating the development of models for a wide range of downstream tasks. Extensive experiments highlight BUSGen's exceptional adaptability, significantly exceeding real-data-trained foundation models in breast cancer screening, diagnosis and prognosis. In breast cancer early diagnosis, our approach outperformed all board-certified radiologists (n = 9), achieving an average sensitivity improvement of 16.5% (P < 0.0001). In addition, we characterized the scaling effect of using synthetic data. Finally, BUSGen enabled de-identified data sharing, making progress forward in secure medical data utilization.

  View details for DOI 10.1038/s41551-026-01639-1

  View details for PubMedID 41946926

  View details for PubMedCentralID 8920425

• Squidiff: predicting cellular development and responses to perturbations using a diffusion model. Nature methods He, S., Zhu, Y., Tavakol, D. N., Ye, H., Lao, Y. H., Zhu, Z., Xu, C., Chauhan, S., Garty, G., Tomer, R., Vunjak-Novakovic, G., Zou, J., Azizi, E., Leong, K. W. 2025

  Abstract

  Single-cell sequencing has revolutionized our understanding of cellular heterogeneity and responses to environmental stimuli. However, mapping transcriptomic changes across diverse cell types in response to various stimuli and elucidating underlying disease mechanisms remains challenging. Here we present Squidiff, a diffusion model-based generative framework that predicts transcriptomic changes across diverse cell types in response to environmental changes. We demonstrate the robustness of Squidiff across cell differentiation, gene perturbation and drug response prediction. Through continuous denoising and semantic feature integration, Squidiff learns transient cell states and predicts high-resolution transcriptomic landscapes over time and conditions. Furthermore, we applied Squidiff to model blood vessel organoid development and cellular responses to neutron irradiation and growth factors. Our results demonstrate that Squidiff enables in silico screening of molecular landscapes and cellular state transitions, facilitating rapid hypothesis generation and providing valuable insights into the regulatory principles of cell fate decisions.

  View details for DOI 10.1038/s41592-025-02877-y

  View details for PubMedID 41184550

  View details for PubMedCentralID 5823237

• Predicting cellular responses with conditional diffusion models NATURE METHODS He, S., Leong, K. W. 2025

  View details for DOI 10.1038/s41592-025-02878-x

  View details for Web of Science ID 001607947500001

  View details for PubMedID 41184557

  View details for PubMedCentralID 12148494

• Benchmarking cell type and gene set annotation by large language models with AnnDictionary. Nature communications Crowley, G., Quake, S. R. 2025; 16 (1): 9511

  Abstract

  We develop an open-source package called AnnDictionary to facilitate the parallel, independent analysis of multiple anndata. AnnDictionary is built on top of LangChain and AnnData and supports all common large language model (LLM) providers. AnnDictionary only requires 1 line of code to configure or switch the LLM backend and it contains numerous multithreading optimizations to support the analysis of many anndata and large anndata. We use AnnDictionary to perform the first benchmarking study of all major LLMs at de novo cell-type annotation. LLMs vary greatly in absolute agreement with manual annotation based on model size. Inter-LLM agreement also varies with model size. We find that LLM annotation of most major cell types to be more than 80-90% accurate, and will maintain a leaderboard of LLM cell type annotation. Furthermore, we benchmark these LLMs at functional annotation of gene sets, and find that Claude 3.5 Sonnet recovers close matches of functional gene set annotations in over 80% of test sets.

  View details for DOI 10.1038/s41467-025-64511-x

  View details for PubMedID 41152246

  View details for PubMedCentralID 8080633

• Squidiff: Predicting cellular development and responses to perturbations using a diffusion model. bioRxiv : the preprint server for biology He, S., Zhu, Y., Tavakol, D. N., Ye, H., Lao, Y. H., Zhu, Z., Xu, C., Chauhan, S., Garty, G., Tomer, R., Vunjak-Novakovic, G., Zou, J., Azizi, E., Leong, K. W. 2025

  Abstract

  Single-cell sequencing has revolutionized our understanding of cellular heterogeneity and responses to environmental stimuli. However, mapping transcriptomic changes across diverse cell types in response to various stimuli and elucidating underlying disease mechanisms remains challenging. Studies involving physical stimuli, such as radiotherapy, or chemical stimuli, like drug testing, demand labor-intensive experimentation, hindering mechanistic insight and drug discovery. Here we present Squidiff, a diffusion model-based generative framework that predicts transcriptomic changes across diverse cell types in response to environmental changes. We demonstrate Squidiff's robustness across cell differentiation, gene perturbation, and drug response prediction. Through continuous denoising and semantic feature integration, Squidiff learns transient cell states and predicts high-resolution transcriptomic landscapes over time and conditions. Furthermore, we applied Squidiff to model blood vessel organoid development and cellular responses to neutron irradiation and growth factors. Our results demonstrate that Squidiff enables in silico screening of molecular landscapes, facilitating rapid hypothesis generation and providing valuable insights for precision medicine.

  View details for DOI 10.1101/2024.11.16.623974

  View details for PubMedID 40909548

  View details for PubMedCentralID PMC12407682

• Quantifying large language model usage in scientific papers. Nature human behaviour Liang, W., Zhang, Y., Wu, Z., Lepp, H., Ji, W., Zhao, X., Cao, H., Liu, S., He, S., Huang, Z., Yang, D., Potts, C., Manning, C. D., Zou, J. 2025

  Abstract

  Scientific publishing is the primary means of disseminating research findings. There has been speculation about how extensively large language models (LLMs) are being used in academic writing. Here we conduct a systematic analysis across 1,121,912 preprints and published papers from January 2020 to September 2024 on arXiv, bioRxiv and Nature portfolio journals, using a population-level framework based on word frequency shifts to estimate the prevalence of LLM-modified content over time. Our findings suggest a steady increase in LLM usage, with the largest and fastest growth estimated for computer science papers (up to 22%). By comparison, mathematics papers and the Nature portfolio showed lower evidence of LLM modification (up to 9%). LLM modification estimates were higher among papers from first authors who post preprints more frequently, papers in more crowded research areas and papers of shorter lengths. Our findings suggest that LLMs are being broadly used in scientific writing.

  View details for DOI 10.1038/s41562-025-02273-8

  View details for PubMedID 40760036

  View details for PubMedCentralID 5199034

• Spatial multi-omics and deep learning reveal fingerprints of immunotherapy response and resistance in hepatocellular carcinoma. bioRxiv : the preprint server for biology Wu, Z., Boen, J., Jindal, S., Basu, S., Bieniosek, M., He, S., LaPelusa, M., Mayer, A. T., Kaseb, A. O., Zou, J., Sharma, P., Trevino, A. E. 2025

  Abstract

  Despite advances in immunotherapy treatment, nonresponse rates remain high, and mechanisms of resistance to checkpoint inhibition remain unclear. To address this gap, we performed spatial transcriptomic and proteomic profiling on human hepatocellular carcinoma tissues collected before and after immunotherapy. We developed an interpretable, multimodal deep learning framework to extract key cellular and molecular signatures from these data. Our graph neural network approach based on spatial proteomic inputs achieved outstanding performance (ROC-AUC > 0.9) in predicting patient treatment response. Key predictive features and associated spatial transcriptomic profiles revealed the multi-omic landscape of immunotherapy response and resistance. One such feature was an interface niche expressing restrictive extracellular matrix factors that physically separates tumor tissue and lymphoid aggregates in nonresponders. We integrate this and other spatially-resolved signatures into SPARC, a multi-omic "fingerprint" comprising scores for immunotherapy response and resistance mechanisms. This study lays groundwork for future patient stratification and treatment strategies in cancer immunotherapy.

  View details for DOI 10.1101/2025.06.11.656869

  View details for PubMedID 40661489

  View details for PubMedCentralID PMC12259099

• Encoding spatial tumour dynamics with Starfysh NATURE REVIEWS CANCER He, S. 2024

  View details for DOI 10.1038/s41568-024-00764-w

  View details for Web of Science ID 001330068300001

  View details for PubMedID 39394485

• Starfysh integrates spatial transcriptomic and histologic data to reveal heterogeneous tumor-immune hubs NATURE BIOTECHNOLOGY He, S., Jin, Y., Nazaret, A., Shi, L., Chen, X., Rampersaud, S., Dhillon, B. S., Valdez, I., Friend, L. E., Fan, J., Park, C. Y., Mintz, R. L., Lao, Y., Carrera, D., Fang, K. W., Mehdi, K., Rohde, M., McFaline-Figueroa, J. L., Blei, D., Leong, K. W., Rudensky, A. Y., Plitas, G., Azizi, E. 2024

  Abstract

  Spatially resolved gene expression profiling provides insight into tissue organization and cell-cell crosstalk; however, sequencing-based spatial transcriptomics (ST) lacks single-cell resolution. Current ST analysis methods require single-cell RNA sequencing data as a reference for rigorous interpretation of cell states, mostly do not use associated histology images and are not capable of inferring shared neighborhoods across multiple tissues. Here we present Starfysh, a computational toolbox using a deep generative model that incorporates archetypal analysis and any known cell type markers to characterize known or new tissue-specific cell states without a single-cell reference. Starfysh improves the characterization of spatial dynamics in complex tissues using histology images and enables the comparison of niches as spatial hubs across tissues. Integrative analysis of primary estrogen receptor (ER)-positive breast cancer, triple-negative breast cancer (TNBC) and metaplastic breast cancer (MBC) tissues led to the identification of spatial hubs with patient- and disease-specific cell type compositions and revealed metabolic reprogramming shaping immunosuppressive hubs in aggressive MBC.

  View details for DOI 10.1038/s41587-024-02173-8

  View details for Web of Science ID 001190085400001

  View details for PubMedID 38514799

  View details for PubMedCentralID 9118175

• Human vascular organoids with a mosaicAKT1mutation recapitulate Proteus syndrome. bioRxiv : the preprint server for biology He, S., Zhu, Y., Chauhan, S., Tavakol, D. N., Lee, J. H., Berris, R. B., Xu, C., Lee, J. H., Lee, C., Cai, S., McElroy, S., Vunjak-Novakovic, G., Tomer, R., Azizi, E., Xu, B., Lao, Y., Leong, K. W. 2024

  Abstract

  Vascular malformation, a key clinical phenotype of Proteus syndrome, lacks effective models for pathophysiological study and drug development due to limited patient sample access. To bridge this gap, we built a human vascular organoid model replicating Proteus syndrome's vasculature. Using CRISPR/Cas9 genome editing and gene overexpression, we created induced pluripotent stem cells (iPSCs) embodying the Proteus syndrome-specific AKTE17K point mutation for organoid generation. Our findings revealed that AKT overactivation in these organoids resulted in smaller sizes yet increased vascular connectivity, although with less stable connections. This could be due to the significant vasculogenesis induced by AKT overactivation. This phenomenon likely stems from boosted vasculogenesis triggered by AKT overactivation, leading to increased vascular sprouting. Additionally, a notable increase in dysfunctional PDGFRbeta+ mural cells, impaired in matrix secretion, was observed in these AKT-overactivated organoids. The application of AKT inhibitors (ARQ092, AZD5363, or GDC0068) reversed the vascular malformations; the inhibitors' effectiveness was directly linked to reduced connectivity in the organoids. In summary, our study introduces an innovative in vitro model combining organoid technology and gene editing to explore vascular pathophysiology in Proteus syndrome. This model not only simulates Proteus syndrome vasculature but also holds potential for mimicking vasculatures of other genetically driven diseases. It represents an advance in drug development for rare diseases, historically plagued by slow progress.

  View details for DOI 10.1101/2024.01.26.577324

  View details for PubMedID 38328122

• Emulating Early Atherosclerosis in a Vascular Microphysiological System Using Branched Tissue-Engineered Blood Vessels ADVANCED BIOLOGY Lee, J. H., Chen, Z., He, S., Zhou, J. K., Tsai, A., Truskey, G. A., Leong, K. W. 2021; 5 (4): e2000428

  Abstract

  Atherosclerosis begins with the accumulation of cholesterol-carrying lipoproteins on blood vessel walls and progresses to endothelial cell dysfunction, monocyte adhesion, and foam cell formation. Endothelialized tissue-engineered blood vessels (TEBVs) have previously been fabricated to recapitulate artery functionalities, including vasoconstriction, vasodilation, and endothelium activation. Here, the initiation of atherosclerosis is emulated by designing branched TEBVs (brTEBVs) of various geometries treated with enzyme-modified low-density-lipoprotein (eLDL) and TNF-α to induce endothelial cell dysfunction and adhesion of perfused human monocytes. Locations of monocyte adhesion under pulsatile flow are identified, and the hemodynamics in the brTEBVs are characterized using particle image velocimetry (PIV) and computational fluid dynamics (CFD). Monocyte adhesion is greater at the side outlets than at the main outlets or inlets, and is greatest at larger side outlet branching angles (60° or 80° vs 45°). In PIV experiments, the branched side outlets are identified as atherosclerosis-prone areas where fluorescent particles show a transient swirling motion following flow pulses; in CFD simulations, side outlets with larger branching angles show higher vorticity magnitude and greater flow disturbance than other areas. These results suggest that the branched TEBVs with eLDL/TNF-α treatment provide a physiologically relevant model of early atherosclerosis for preclinical studies.

  View details for DOI 10.1002/adbi.202000428

  View details for Web of Science ID 000672525100006

  View details for PubMedID 33852179

  View details for PubMedCentralID PMC9951769

• Point-of-care detection, characterization, and removal of chocolate bloom using a handheld Raman spectrometer SCIENTIFIC REPORTS Heuler, J., He, S., Ambardar, S., Voronine, D. V. 2020; 10 (1): 9833

  Abstract

  Chocolate bloom is an off-white coating on the surface of chocolate products due to the altered distribution of the ingredients. Bloom reduces the shelf-life of chocolate and affects its visual and tactile quality, all of which are serious concerns for chocolate manufacturers and consumers. The automated, rapid, and noninvasive point-of-care detection of chocolate bloom has been an essential but challenging problem. The ability to detect and characterize chocolate bloom using portable laser spectroscopy could be used to develop in-situ quality control sensors. In this work, a handheld Raman spectrometer was used to detect chocolate bloom. Raman spectra acquired from bloomed HERSHEY'S milk chocolate, Hawaiian Host milk chocolate covered macadamia nuts, and Babayevsky Russian dark chocolate were used to characterize the type of bloom. The 1064 nm laser beam of the handheld Raman instrument was used to partially remove the fat bloom of the dark chocolate and to induce sugar bloom on the milk chocolate. The handheld Raman approach has a high potential for industrial and consumer applications for the on-site chemical analysis of chocolate bloom and as an alternative laser-based chocolate decoration.

  View details for DOI 10.1038/s41598-020-66820-1

  View details for Web of Science ID 000543958300002

  View details for PubMedID 32555247

  View details for PubMedCentralID PMC7299933

• CRISPR/Cas9-mediated mutagenesis to validate the synergy between PARP1 inhibition and chemotherapy in <i>BRCA1</i>-mutated breast cancer cells BIOENGINEERING & TRANSLATIONAL MEDICINE Mintz, R. L., Lao, Y., Chi, C., He, S., Li, M., Quek, C., Shao, D., Chen, B., Han, J., Wang, S., Leong, K. W. 2020; 5 (1): e10152

  Abstract

  For patients carrying BRCA1 mutations, at least one-third develop triple negative breast cancer (TNBC). Not only is TNBC difficult to treat due to the lack of molecular target receptors, but BRCA1 mutations (BRCA1m) also result in chemotherapeutic resistance, making disease recurrence more likely. Although BRCA1m are highly heterogeneous and therefore difficult to target, BRCA1 gene's synthetic lethal pair, PARP1, is conserved in BRCA1m cancer cells. Therefore, we hypothesize that targeting PARP1 might be a fruitful direction to sensitize BRCA1m cancer cells to chemotherapy. We used CRISPR/Cas9 technology to generate PARP1 deficiency in two TNBC cell lines, MDA-MB-231 (BRCA1 wild-type) and MDA-MB-436 (BRCA1m). We explored whether this PARP1 disruption (PARP1m) could significantly lower the chemotherapeutic dose necessary to achieve therapeutic efficacy in both a 2D and 3D tumor-on-a-chip model. With both BRCA1m and PARP1m, the TNBC cells were more sensitive to three representative chemotherapeutic breast cancer drugs, doxorubicin, gemcitabine and docetaxel, compared with the PARP1 wild-type counterpart in the 2D culture environment. However, PARP1m did not result in this synergy in the 3D tumor-on-a-chip model, suggesting that drug dosing in the tumor microenvironment may influence the synergy. Taken together, our results highlight a discrepancy in the efficacy of the combination of PARP1 inhibition and chemotherapy for TNBC treatment, which should be clarified to justify further clinical testing.

  View details for DOI 10.1002/btm2.10152

  View details for Web of Science ID 000505073400001

  View details for PubMedID 31989039

  View details for PubMedCentralID PMC6971465

• Tip-enhanced Raman scattering of DNA aptamers for <i>Listeria monocytogenes</i> BIOINTERPHASES He, S., Li, H., Gomes, C. L., Voronine, D. V. 2018; 13 (3): 03C402

  Abstract

  Optical detection and conformational mapping of aptamers are important for improving medical and biosensing technologies and for better understanding of biological processes at the molecular level. The authors investigate the vibrational signals of deoxyribonucleic acid aptamers specific to Listeria monocytogenes immobilized on gold substrates using tip-enhanced Raman scattering (TERS) spectroscopy and nanoscale imaging. The authors compare topographic and nano-optical signals and investigate the fluctuations of the position-dependent TERS spectra. They perform spatial TERS mapping with 3 nm step size and discuss the limitation of the resulting spatial resolution under the ambient conditions. TERS mapping provides information about the chemical composition and conformation of aptamers and paves the way to future label-free biosensing.

  View details for DOI 10.1116/1.5022303

  View details for Web of Science ID 000431347700001

  View details for PubMedID 29724106

• Spontaneous formation of non-uniform double helices for elastic rods under torsion PHYSICS LETTERS A Li, H., Zhao, S., Xia, M., He, S., Yang, Q., Yan, Y., Zhao, H. 2017; 381 (7): 689-700

  View details for DOI 10.1016/j.physleta.2016.12.008

  View details for Web of Science ID 000392678100002

• OBSERVATIONAL CONSTRAINTS ON FIRST-STAR NUCLEOSYNTHESIS. I . EVIDENCE FOR MULTIPLE PROGENITORS OF CEMP-NO STARS ASTROPHYSICAL JOURNAL Yoon, J., Beers, T. C., Placco, V. M., Rasmussen, K. C., Carollo, D., He, S., Hansen, T. T., Roederer, I. U., Zeanah, J. 2016; 833 (1)

  View details for DOI 10.3847/0004-637X/833/1/20

  View details for Web of Science ID 000390326600001