About

Rui Kuang is a professor in the Department of Computer Science & Engineering at the University of Minnesota, joined in 2006 and promoted to full professor in 2021. His research interests encompass computational biology, biomedical informatics, and machine learning. His lab focuses on developing machine-learning algorithms for problems in cancer genomics, biological network analysis, and protein function and structure analysis. Kuang holds a Ph.D. in Computer Science from Columbia University, an M.S. from Temple University, and a B.S. from Nankai University. Prior to his current position, he served as a graduate assistant at Columbia University and Temple University, and worked as a system analyst for Procter & Gamble in China. He has received notable recognition including the NSF CAREER award in 2012. Kuang also serves as an adjunct professor of bioinformatics and computational biology.

Research topics

• Computer Science
• Artificial Intelligence
• Engineering
• Electronic engineering
• Machine Learning
• Electrical engineering

Selected publications

• Cytoplasmic HMGB1 promotes and interacts with BECN1 through ZNF460 to induce autophagy and accelerate radioresistance in colorectal cancer cells

  Frontiers in Immunology · 2025-10-14 · 2 citations

  articleOpen access

  Radioresistance results in relapse and treatment failure in locally advanced colorectal cancer (CRC) patients. HMGB1 is reportedly associated with radioresistance in esophageal squamous cell carcinoma and breast cancer. However, its role in the response of CRC to radiotherapy has not been fully elucidated. Thus, we explored the role and underlying mechanism of HMGB1 in CRC radioresistance. The total amount of HMGB1 and its translocation from the nucleus to the cytoplasm increased after irradiation. Functional studies revealed that HMGB1 enhanced the proliferation and autophagy of CRC cells after irradiation. Mechanistically, HMGB1 can regulate the transcription factor ZNF460, which combines with the BECN1 promoter to promote the release of BECN1 into the cytoplasm after irradiation. Moreover, HMGB1 directly interacts with BECN1 in the cytoplasm, thereby resulting in CRC radioresistance. Finally, the protein expression levels of BECN1, which was positively correlated with HMGB1, were significantly increased in human CRC tissues and associated with TNM stage and poor prognosis in patients with CRC. Our findings revealed that HMGB1 plays a vital role in CRC radioresistance by regulating autophagy through binding with BECN1. Given the efficacy of HMGB1 modulation in CRC suppression and radioresistance, HMGB1 has emerged as a potential therapeutic molecule for CRC treatment.

  PublisherOA PDFDOI

• The role of autophagy in spinal cord injury: Mechanisms, crosstalk, and therapeutic strategies

  Neural Regeneration Research · 2025-05-06

  articleOpen access

  Spinal cord injury is a neurological disorder resulting from trauma, typically affecting sensory and motor function at the injury site, even leading to paralysis and internal dysfunction. The treatment of spinal cord injury mainly relies on pharmacological and surgical interventions; however, significant challenges remain in the protection and repair of neural tissues. Autophagy, an intracellular process responsible for the degradation and recycling of macromolecular components, plays a vital role in spinal cord injury, alleviating the severity of injury by inhibiting cell apoptosis and inflammatory responses. In this review, we provide an overview of the physiological mechanisms underlying autophagy and spinal cord injury and detail the crosstalk between autophagy and other modes of cell death in spinal cord injury. In addition, we discuss the potential of targeting autophagy as a therapeutic strategy for spinal cord injury through approaches that focus on promoting or inhibiting this process, targeting specific autophagic substrates or pathways, and combining autophagy modulation with other neuroprotective or restorative interventions. In summary, this review proposes that strict regulation of autophagy may represent a viable strategy for the treatment of spinal cord injury.

  PublisherDOI

• Topological entropy dimensions of free semigroup actions for any subsets

  Discrete and Continuous Dynamical Systems - S · 2025-04-23 · 1 citations

  articleOpen access

  Similar to the entropy dimensions of dynamical systems formed by a single map, we introduce the concept of entropy dimension of free semigroup actions for any set to classify sets with entropy zero. We compare the entropy dimensions induced by different topological entropies, providing examples to demonstrate their distinctions. Moreover, we present their basic properties as follows. A class of free semigroup actions with the entropy dimension 0 is identified. Countable stability and power rule for entropy dimension are established. It is proved that the entropy dimension of a free semigroup action is smaller than the entropy dimension of a system formed by any one of its generators. For any free semigroup action, we show that the entropy dimension of any set is equal to the largest entropy dimension among images of the set under generators of the free semigroup action. Furthermore, the entropy dimension of free semigroup actions is invariant under conjugacy.

  PublisherOA PDFDOI

• A ventriculomegaly feature computational pipeline to improve the screening of normal pressure hydrocephalus on CT

  Journal of neurosurgery · 2024-03-09 · 6 citations

  article

  OBJECTIVE: The objective of this study was to develop a computational pipeline that extracts objective features of ventriculomegaly from non-contrast CT (NCCT) for the accurate classification of idiopathic normal pressure hydrocephalus (NPH) from headache controls (HCs), Alzheimer's dementia (AD), and posttraumatic encephalomalacia (PTE). METHODS: Patients with possible NPH (n = 79) and a subset with definite NPH (DefNPH; n = 29) were retrospectively identified in the Veterans Affairs Informatics and Computing Infrastructure system, along with the AD (n = 62), PTE (n = 53), and HC (n = 59) cohorts. Image-processing pipelines were developed to extract a novel feature capturing the maximum eccentricity of the lateral ventricles (MaxEccLV), a proxy splenial angle (p-SA), the Evans indices (EI-x, -y, and -z), callosal angle, normalized maximum third-ventricle width, and CSF to brain volume ratio from their NCCT scans. The authors used t-tests to examine group differences in the features and multivariate logistic regression models for classification. Additionally, the NPH versus HC classifier was validated on external data. RESULTS: When NPH and DefNPH were compared with HC, AD, and PTE, significant differences were found in all features except the p-SA, which only significantly differed between NPH and PTE. The test-set area under the receiver operating characteristic curve (AUC), sensitivity, and specificity were 0.98, 100%, and 98.3% for NPH versus HC classification; 0.94, 87.3%, and 85.5% for NPH versus AD; 0.96, 92.4%, and 90.6% for NPH versus PTE; and 0.96, 94%, and 88% for NPH versus the other groups using logistic regression under five-fold cross-validation. Consistently high performance was noted for DefNPH. The NPH versus HC classifier provided an AUC of 0.84, sensitivity of 76.9%, and specificity of 90% when assessed on external data. CONCLUSIONS: Including the novel MaxEccLV, this framework computes useful features of ventriculomegaly, which had not previously been algorithmically assessed on NCCT. This framework successfully classified possible and definite NPH from HC, AD, and PTE. Following validation on larger representative cohorts, this objective and accessible tool may aid in screening for NPH and differentiating it from symptomatic mimics such as AD and PTE.

  PublisherDOI

• Structural neuroimaging markers of normal pressure hydrocephalus versus Alzheimer’s dementia and Parkinson’s disease, and hydrocephalus versus atrophy in chronic TBI—a narrative review

  Frontiers in Neurology · 2024-03-21 · 7 citations

  articleOpen access

  Introduction: Normal Pressure Hydrocephalus (NPH) is a prominent type of reversible dementia that may be treated with shunt surgery, and it is crucial to differentiate it from irreversible degeneration caused by its symptomatic mimics like Alzheimer's Dementia (AD) and Parkinson's Disease (PD). Similarly, it is important to distinguish between (normal pressure) hydrocephalus and irreversible atrophy/degeneration which are among the chronic effects of Traumatic Brain Injury (cTBI), as the former may be reversed through shunt placement. The purpose of this review is to elucidate the structural imaging markers which may be foundational to the development of accurate, noninvasive, and accessible solutions to this problem. Methods: By searching the PubMed database for keywords related to NPH, AD, PD, and cTBI, we reviewed studies that examined the (1) distinct neuroanatomical markers of degeneration in NPH versus AD and PD, and atrophy versus hydrocephalus in cTBI and (2) computational methods for their (semi-) automatic assessment on Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) scans. Results: Structural markers of NPH and those that can distinguish it from AD have been well studied, but only a few studies have explored its structural distinction between PD. The structural implications of cTBI over time have been studied. But neuroanatomical markers that can predict shunt response in patients with either symptomatic idiopathic NPH or post-traumatic hydrocephalus have not been reliably established. MRI-based markers dominate this field of investigation as compared to CT, which is also reflected in the disproportionate number of MRI-based computational methods for their automatic assessment. Conclusion: Along with an up-to-date literature review on the structural neurodegeneration due to NPH versus AD/PD, and hydrocephalus versus atrophy in cTBI, this article sheds light on the potential of structural imaging markers as (differential) diagnostic aids for the timely recognition of patients with reversible (normal pressure) hydrocephalus, and opportunities to develop computational tools for their objective assessment.

  PublisherOA PDFDOI

• Free vibration of the piezoelectric laminated composite beams under the elastic foundation

  Physica Scripta · 2024-03-13 · 2 citations

  articleOpen access

  Abstract Due to its self-adaptability and electric-mechanical interaction, the piezoelectric laminated composite structure has received increased attention in the engineering community, with the intricacy of the actual scenario, the structural stability based on numerous foundations has also been recognized. In this paper, a unified formulation which is based on classical laminated beam is presented to conduct free vibration analysis of piezoelectric laminated beams subjected to classical boundary conditions with the varied elastic foundations. The semi-analytical solutions of layered beams in various boundaries have been obtained based on the two-dimensional piezoelectric elastic theory and state space-differential quadrature method (SS-DQM), the solution of the general structural equation has been generated introducing the electrical boundary conditions and interlayer continuity conditions with this understanding. The different elastic foundations have been used to calculate the frequency equation of the piezoelectric laminated construction. The proposed finite element model is validated through the analysis of the piezoelectric laminated composite beam (PLCB) using the SS-DQM, and the present solutions are compared with those available in the literature to confirm their validity. A systematic parameter study for the piezoelectric laminated beams with various boundary conditions, piezoelectric parameters, lamina numbers and thickness ratios is also performed. The results of the PLCBs with various boundary constraints under the elastic foundation are presented and they may be served as a benchmark for researchers in this field.

  PublisherDOI

• 442 A Ventriculomegaly Feature Computation Pipeline to Improve the Screening of Normal Pressure Hydrocephalus on Computed Tomography

  Neurosurgery · 2024-03-15

  article

  INTRODUCTION: Normal pressure hydrocephalus (NPH) commonly presents with a triad of gait disturbance, incontinence, and memory impairment. Shunt placement effectively treats symptoms; however, NPH is commonly misdiagnosed with below 20% of patients receiving a diagnosis. Given undiagnosed NPH elderly patients often present to the emergency room (ER) with falls, algorithmic identification of NPH using non-contrast CT (NCCT) may improve diagnosis rate. METHODS: NCCTs for 84 patients with NPH, 72 headache controls (HC), 97 Alzheimer’s Disease (AD) patients, and 61 post-traumatic encephalomalacia (PTE) patients were identified using Veterans Affairs Informatics and Computing Infrastructure. Image processing pipelines were developed to extract a novel feature capturing lateral ventricle eccentricity (MaxEccLV), a proxy-Splenial Angle (p-SA), the Evans indices (EI-x, y, z), Callosal angle (CA), third ventricle width (NMax-3VW), and CSF to brain volume ratio (CSF2BVR). T-tests were used to examine group differences and logistic regression to classify between NPH, AD, PTE, and HC. RESULTS: The AUC, sensitivity, and specificity for NPH versus AD+HC+PTE was 0.94, 83.0%, and 87.0% respectively. For NPH versus HC, it was 0.98, 98.8%, and 98.6% respectively. For NPH versus AD, it was 0.90, 84.5%, and 82.5% respectively. For NPH versus PTE, it was 0.95, 91.7%, and 86.9% respectively. Significant differences were seen between NPH and HC, AD, and PTE in MaxEccLV and all other features. CONCLUSIONS: The computational pipeline proposed here can identify NPH from HCs and ventriculomegaly-causing conditions using NCCT with a high sensitivity and specificity. This may help capture undiagnosed NPH patients receiving a NCCT due to presentation to the ER as a fall.

  PublisherDOI

• Acoustoelastic Theory and Mode Analysis of Bolted Structures Under Preload

  Machines · 2024-11-18 · 4 citations

  articleOpen access

  Bolted connections are a common feature of connection in mechanical structures, employed to secure connected parts by tightening nuts and providing preload. The preload is susceptible to various factors leading to potential bolt loosening. The acoustoelastic theory is the most common measure of a bolt structure’s stress. The present study investigates the relationship between the inherent properties of a structure and its acousticelastic properties. The modal response of the bolted structure under different preload forces is studied by translating the acoustoelastic relationship of the structure into an analysis of its intrinsic properties. The modal analysis reflects the relative change in wave velocity to be determined implicitly based on the eigenfrequencies of the structure. A frequency formulation of classical bolted structures based on acoustoelastic theory is presented in this paper to conduct the intrinsic characteristic analysis of bolted structures. The COMSOL5.4 simulation results are under the acoustic elasticity coefficients for ultrasonic wave propagation in bolt structures, as predicted by the acoustic elasticity theory, and the present solutions are compared with those available in the literature to confirm their validity. A systematic parameter study for bolted structures under the varying preloads with different material parameters, Lame elastic constants, Murnaghan third-order elastic constants, and structural parameters are presented. These results may serve as a benchmark for researchers in this field.

  PublisherDOI

• Predicting spatially resolved gene expression via tissue morphology using adaptive spatial GNNs

  Bioinformatics · 2024-06-19 · 6 citations

  articleOpen access

  MOTIVATION: Spatial transcriptomics technologies, which generate a spatial map of gene activity, can deepen the understanding of tissue architecture and its molecular underpinnings in health and disease. However, the high cost makes these technologies difficult to use in practice. Histological images co-registered with targeted tissues are more affordable and routinely generated in many research and clinical studies. Hence, predicting spatial gene expression from the morphological clues embedded in tissue histological images provides a scalable alternative approach to decoding tissue complexity. RESULTS: Here, we present a graph neural network based framework to predict the spatial expression of highly expressed genes from tissue histological images. Extensive experiments on two separate breast cancer data cohorts demonstrate that our method improves the prediction performance compared to the state-of-the-art, and that our model can be used to better delineate spatial domains of biological interest. AVAILABILITY AND IMPLEMENTATION: https://github.com/song0309/asGNN/.

  PublisherOA PDFDOI

• When distribution measures are invariant for dynamical systems

  Monatshefte für Mathematik · 2024-12-20

  article
  PublisherDOI

Recent grants

• CAREER: Predicting and Mining Phenome-genome Association across Species

  NSF · $447k · 2012–2017

• III: Small: Network Learning for Integrative Cancer Genomics

  NSF · $500k · 2011–2015

• IIBR Informatics: Mining Spatial and Single-cell Transcriptomes to Understand Cell Locality and Heterogeneity in Tissues

  NSF · $806k · 2021–2026

Frequent coauthors

• Wei Zhang

  University of Central Florida

  37 shared

• Zhuliu Li

  Twin Cities Orthopedics

  33 shared

• Tianci Song

  University of Minnesota

  31 shared

• Raphael Petegrosso

  Amazon (United States)

  29 shared

• Tae-Hyun Hwang

  21 shared

• Ze Tian

  19 shared

• David Roe

  18 shared

• Jeongsik Yong

  University of Minnesota

  18 shared

Labs

• Kuang LabPI

Awards & honors

• National Science Foundation Faculty Early Career Development…