
Yuchen Cui
· Assistant ProfessorVerifiedUniversity of California, Los Angeles · Computer Science
Active 2016–2025
About
Yuchen Cui is an Assistant Professor in the Department of Computer Science at UCLA Samueli School of Engineering. His research focuses on machine learning and human-robot interaction, particularly for applications in home robotics. Cui's work involves developing methods for distilling and retrieving generalizable knowledge for robot manipulation through language corrections, creating hybrid robot actions for imitation learning, and designing gesture-informed robot assistance using foundation models. He has contributed to advancing the understanding of online language corrections for robotic manipulation via shared autonomy. Cui earned his PhD in Computer Science from the University of Texas at Austin in 2021.
Research topics
- Genetics
- Biology
- Medicine
- Virology
- Cell biology
- Molecular biology
- Computational biology
- Pathology
Selected publications
Orthopedic infections associated with distinct Acinetobacter strains in rural area of Qingdao, China
Frontiers in Cellular and Infection Microbiology · 2025-07-30
articleOpen accessIntroduction Acinetobacter baumannii poses a profound global health threat because of multidrug resistance and its association with nosocomial infections. However, standard clinical diagnostics often report it together with other Acinetobacter species as A. baumannii-calcoaceticus complex (ABC), which unavoidably conceals the attribution of non- A. baumannii species. This study reported orthopedic infection cases associated with different Acinetobacter species and characterized the genomes of the culture isolates to evaluate their potential impact on the clinical treatment. Methods Nine in-patients with A. baumannii-calcoaceticus complex identified by culture during hospitalization were enrolled by the Orthopedics Department from a local hospital in Qingdao, China. Their clinical data were reviewed. One ABC isolate from each patient was tested for drug susceptibility and subjected for whole-genome sequencing, followed by bioinformatic analyses. Results Through whole-genome analysis, nine ABC isolates were identified as six A. baumannii , two A. pittii , and one A. soli with distinct antibiotic resistance profiles and phylogenetic characteristics, indicating progressing pathogen transmission across broad geographic regions in One Health perspective. All A. baumannii and A. pittii strains carried multidrug resistance genes, while A. soli bore only amvA and rsmA . Phenotypically, eight isolates were susceptible to almost all the antibiotics tested, with only one A. baumannii being multidrug resistant. Despite this, eight patients received cephalosporins following positive reports of A. baumannii-calcoaceticus complex. Conclusion Our study highlighted the limitation of current clinical diagnostic approaches for non- A. baumannii cases, which tended to be overtreated, and suggested that Acinetobacter etiology landscape should be explored further beyond A. baumannii to avoid antibiotic misuse.
Viruses · 2025-02-14 · 6 citations
articleOpen accessPorcine epidemic diarrhea virus (PEDV) is a highly contagious pathogen responsible for devastating enteric disease and lethal watery diarrhea, leading to significant economic losses in the global swine industry. Understanding the epidemiology and genetic diversity of PEDV over the past decade is crucial for the effective prevention and treatment of porcine epidemic diarrhea. In this study, 1851 fecal samples were collected from pigs exhibiting diarrhea symptoms across 11 cities in Yunnan Province between 2013 and 2022. The prevalence of PEDV, along with other common swine diarrhea viruses, including porcine transmissible gastroenteritis virus (TGEV), porcine rotavirus (PoRV), porcine Sapporo virus (PoSaV), porcine stellate virus (PaStV), and porcine delta coronavirus (PDCoV) was assessed using a polymerase chain reaction (PCR) assay. The results revealed a total detection rate of 52.94% (980/1851) for the six viruses, with PEDV accounting for 25.93% (480/1851) of cases. Further analysis showed that weaned piglets were more susceptible to PEDV than fattening pigs, with the highest prevalence observed in spring (61.52%, 275/447) and the lowest in summer (12.68%, 97/765). Dual infections were also identified, with PEDV + PoSaV being the most common combination (2.81%, 52/1851), followed by PEDV + PoRV, with a detection rate of 1.67% (31/1851). Phylogenetic analysis of the PEDV S genes revealed that the 28 epidemic strains in Yunnan Province shared a nucleotide sequence homology from 91.4% to 98.4% and an amino acid sequence homology ranging from 85.6% to 99.3%. All strains were classified as GII variant strains. This study provides a comprehensive overview of the epidemiology of PEDV and its co-infection patterns with other common diarrhea-causing viruses in the swine herds of Yunnan Province over the past decade. These findings offer valuable insights for the development of effective prevention and control strategies to mitigate the impact of PEDV and other enteroviruses on the swine industry in Yunnan Province.
Frontiers in Microbiology · 2024-06-07 · 5 citations
articleOpen accessIntroduction Clinical significance of coagulase-negative staphylococci (CoNS) has been gradually acknowledged in both healthcare and clinical research, but approaches for their precise discrimination at the species level remain scarce. The current study aimed to evaluate the association of CoNS with orthopedic infections, where accurate and prompt identification of etiology is crucial for appropriate diagnosis and treatment decision-making. Methods A 16S rRNA-based quantitative PCR (qPCR) assay was developed for the detection of Staphylococcus genus and two panels of 3-plex qPCR assays for further differentiation of six CoNS species with remarkable clinical significance, including S. epidermidis , S. haemolyticus , S. simulans , S. hominis , S. capitis , and S. caprae . All the assays exhibited excellent analytical performance. ΔCq (quantification cycle) between 16S rRNA and CoNS species-specific targets was established to determine the primary CoNS. These methods were applied to detect CoNS in wound samples from orthopedic patients with and without infection. Results and discussion Overall, CoNS were detected in 17.8% (21/118) of patients with clinically suspected infection and in 9.8% (12/123) of patients without any infection symptom ( p < 0.05). Moreover, the association with infection was found to be bacterial quantity dependent. S. epidermidis was identified as the predominant species, followed by S. simulans , S. haemolyticus , and S. hominis . Male sex, open injury, trauma, and lower extremity were determined as risk factors for CoNS infections. CoNS-positive patients had significantly longer hospitalization duration (20 days (15, 33) versus 13 days (7, 22) for Staphylococcus -negative patients, p = 0.003), which could be a considerable burden for healthcare and individual patients. Considering the complex characteristics and devastating consequences of orthopedic infections, further expanding the detection scope for CoNS may be pursued to better understand the etiology of orthopedic infections and to improve therapeutic strategies.
Viruses · 2024-07-20 · 10 citations
articleOpen accessRetroviral integration is mediated by intasome nucleoprotein complexes wherein a pair of viral DNA ends are bridged together by a multimer of integrase (IN). Atomic-resolution structures of HIV-1 intasomes provide detailed insights into the mechanism of integration and inhibition by clinical IN inhibitors. However, previously described HIV-1 intasomes are highly heterogeneous and have the tendency to form stacks, which is a limiting factor in determining high-resolution cryo-EM maps. We have assembled HIV-1 intasomes in the presence of excess IN C-terminal domain protein, which was readily incorporated into the intasomes. The purified intasomes were largely homogeneous and exhibited minimal stacking tendencies. The cryo-EM map resolution was further improved to 2.01 Å, which will greatly facilitate structural studies of IN inhibitor action and drug resistance mechanisms. The C-terminal 18 residues of HIV-1 IN, which are critical for virus replication and integration in vitro, have not been well resolved in previous intasome structures, and its function remains unclear. We show that the C-terminal tail participates in intasome assembly, resides within the intasome core, and forms a small alpha helix (residues 271-276). Mutations that disrupt alpha helix integrity impede IN activity in vitro and disrupt HIV-1 infection at the step of viral DNA integration.
Frontiers in Cellular and Infection Microbiology · 2024-06-13 · 2 citations
articleOpen accessIntroduction Accurate identification of the etiology of orthopedic infection is very important for correct and timely clinical management, but it has been poorly studied. In the current study we explored the association of multiple bacterial pathogens with orthopedic infection. Methods Hospitalized orthopedic patients were enrolled in a rural hospital in Qingdao, China. Wound or exudate swab samples were collected and tested for twelve bacterial pathogens with both culture and multiplex real time PCR. Results and discussion A total of 349 hospitalized orthopedic patients were enrolled including 193 cases presenting infection manifestations upon admission and 156 with no sign of infection. Orthopedic infection patients were mainly male (72.5%) with more lengthy hospital stay (median 15 days). At least one pathogen was detected in 42.5% (82/193) of patients with infection while 7.1% (11/156) in the patients without infection ( P < 0.001). S. aureus was the most prevalent causative pathogen (15.5%). Quantity dependent pathogen association with infection was observed, particularly for P. aeruginosa and K. pneumoniae , possibly indicating subclinical infection. Most of the patients with detected pathogens had a previous history of orthopedic surgery (odds ratio 2.8, P = 0.038). Pathogen specific clinical manifestations were characterized. Multiplex qPCR, because of its high sensitivity, superior specificity, and powerful quantification could be utilized in combination with culture to guide antimicrobial therapy and track the progression of orthopedic infection during treatment.
Lesion recognition by XPC, TFIIH and XPA in DNA excision repair
Acta Crystallographica Section A Foundations and Advances · 2023-07-07 · 1 citations
articleOpen accessLesion recognition by XPC, TFIIH and XPA in DNA excision repair
Nature · 2023 · 81 citations
- Biology
- Molecular biology
- Genetics
Protein & Cell · 2023-02-04 · 3 citations
articleOpen accessHKIE Transactions · 2022-12-12 · 1 citations
articleOpen accessTo achieve carbon peaking and carbon neutrality targets of mainland China and Hong Kong, carbon reduction in the building sector is a critical task due to its high contribution to total carbon emissions. Decarbonisation approaches in both operation and construction phases should be considered. A building information modelling (BIM)-based integrated design process (IDP) is introduced to connect the design, construction and operation as an entirety from the initial stage of a project to enhance the green building and decarbonisation design. Taking a real case of a Design-Build-Operate project, O•PARK2, this study investigates the benefits and problems in a BIM-based IDP in regard to green building and its potential impact on carbon neutrality during construction. A BIM-based IDP enables decision-making to be based on comprehensive and optimal information, which is well proved by the O•PARK2 project with the BEAM Plus being upgraded from Gold to Platinum. The carbon reduction potential during the construction phase is as high as 45%, which can only be actualised if decarbonisation specialists are involved from the planning phase. Although the BIM-based IDP possesses many unique advantages, the management of its long feedback loop and the complicated procedures should be optimised in the future.
bioRxiv (Cold Spring Harbor Laboratory) · 2022-08-31 · 1 citations
preprintOpen accessAbstract The Reoviridae family of dsRNA viruses is characterized by its members’ capacity for endogenous transcription of their multipartite genomes within proteinaceous capsids of 1 to 3 layers. These viruses share inner core particles (ICPs) that conform to icosahedral, T=2*, symmetry, but differ in two major respects: first, the presence or absence of RNA-capping turrets at each icosahedral vertex; second, the number of additional host-specific capsid layers that are often lost upon cell entry. While the role of these additional layers in host infection is generally understood, the absence of asymmetric ICP structures from turreted, multilayered reoviruses has obfuscated our understanding of how successive removal of these external layers impact the structural organization of the ICP and transcription initiation. Here, we present the 3.3 Å resolution structure of the aquareovirus (ARV) ICP, and atomic models of the capsid proteins VP3 and VP6, transcriptional enzymatic complex (TEC) subunits VP2 and VP4, and RNA-capping turret protein VP1. These structures reveal significant differences when compared to those of the coated ARV, as well as their counterparts in single-layered cytoplasmic polyhedrosis virus (CPV). Compared to the double-layered ARV virion and infectious subvirion particle structures, the ARV ICP undergoes significant capsid expansion and widening of the nucleotide processing channels in its TEC and turret. Thus, the loss of outer capsid layers may regulate transcription initiation in ARV, unlike CPV which relies solely on allosteric regulation by binding transcriptional cofactors. These results shed new light on the mechanism of transcription initiation amongst turreted, multilayered members of Reoviridae .
Frequent coauthors
- 75 shared
Z. Hong Zhou
University of California System
- 16 shared
Xuemin Chen
- 16 shared
Zihang Li
- 16 shared
Kang Zhou
Lanzhou Petrochemical Polytechnic
- 16 shared
Wei Yang
Wuhan University
- 16 shared
Guo‐Qiang Bi
University of Science and Technology of China
- 15 shared
Huaibin Wang
National Institutes of Health
- 10 shared
Xue Yang
Osaka University
- Resume-aware match score
- Save to shortlist
- AI-drafted outreach
See your match with Yuchen Cui
PhdFit ranks faculty by your research interests, methods, and publications — grounded in their actual work, not templates.
- Free to start
- No credit card
- 30-second signup