Resume-aware faculty matching

Find professors who actually fit you

Upload your resume. Four AI agents analyze your background, rank the faculty who fit, inspect their recent research, and help you draft outreach — grounded in their actual work, not templates.

Free to startNo credit cardCancel anytime
Top matches Balanced preset
Dr. Sarah Chen
Stanford · Interpretability · NLP
91
Dr. Marcus Holloway
MIT · Robotics · RL
84
Dr. Aisha Okonkwo
CMU · Fairness · HCI
82
Nova · Professor Researcher · re-ranking top 20…
Chunyu Liu

Chunyu Liu

· PhD ProfessorVerified

Boston University · Biostatistics

Active 1982–2026

h-index95
Citations54.2k
Papers759399 last 5y
Funding$36.6M
See your match with Chunyu Liu — sign in to PhdFit.Sign in

About

Chunyu Liu, PhD, is a Professor of Biostatistics at Boston University School of Public Health. She earned her PhD in Biostatistics with a specialization in statistical genetics from Boston University. Her professional experience includes working as a statistical geneticist at Biogen Idec Inc. in Cambridge, Massachusetts, and contributing to the Population Sciences Branch & Framingham Heart Study (FHS) at NHLBI in Framingham, Massachusetts. Her research focuses on assessing risk factors associated with cardiovascular disease (CVD) and Alzheimer’s disease (AD). Dr. Liu has held leadership roles as the Principal Investigator for multiple R01 and R21 grants since 2018, and she is involved in collaborative research as a lead biostatistician, contributing to over 135 peer-reviewed manuscripts. She is also active in the scientific community as a member of the NIH’s Genetics of Health and Disease Study Section, an Associate Editor for the Cardiovascular Digital Health Journal, and an Academic Editor for PLOS ONE. Her work includes developing analytical methods for mitochondrial DNA sequence variations and participating in various working groups within genetic epidemiology research.

Research topics

  • Biology
  • Genetics
  • Neuroscience
  • Political Science
  • Medicine
  • Computer Science
  • Psychiatry
  • Psychology
  • Nanotechnology
  • Evolutionary biology
  • Photochemistry
  • Nursing
  • Family medicine
  • Computational biology
  • Chemistry
  • Organic chemistry
  • Optoelectronics
  • Medical emergency
  • Materials science
  • Optics

Selected publications

  • Tempol Protects Against Radiation Injury in the Submandibular Gland Through the β‐Catenin Signaling Pathway

    Journal of Oral Pathology and Medicine · 2026-05-14

    articleOpen access

    BACKGROUND: Radiation damage severely impacts salivary gland function and cell survival, especially in head and neck radiotherapy. Tempol (TPL), a free radical scavenger, has shown protective effects against radiation damage. This research aimed to investigate the protective effects of TPL on radiation-induced damage in Hs917.T cells and the submandibular gland (SMG) of C57BL/6 mice, along with the mechanisms involved. METHODS: Human parotid fibroblasts (Hs917.T) were pre-treated with TPL and exposed to ionizing radiation (IR). Protective effects were evaluated using MTT, clonogenic survival assays, flow cytometry, and intracellular reactive oxygen species levels. In vivo, C57BL/6 mice were pre-treated with TPL (275 mg/kg) and exposed to 15 gray (Gy). Effects were assessed by survival rates, body weight changes, histological analysis, and TUNEL staining. Changes in apoptosis-related markers and β-catenin signaling pathway were analyzed, and the role of TPL was verified using the β-catenin inhibitor XAV939. RESULTS: TPL pre-treatment increased cell survival, reduced apoptosis, alleviated cell cycle arrest, and decreased intracellular superoxide and hydrogen peroxide levels in Hs917. T cells. In C57BL/6 mice, pre-treatment with TPL improved survival, mitigated weight loss, reduced SMG damage, and decreased apoptosis. TPL inhibited IR-induced apoptosis by increasing Bcl-2 expression and decreasing Bax and caspase-9 levels. TPL exerted anti-apoptotic and protective effects by upregulating the expression of β-catenin, promoting its nuclear translocation, and inhibiting its phosphorylation. These protective effects of TPL were reversed by XAV939. CONCLUSIONS: TPL exerted protective effects against IR-induced damage in Hs917.T cells and the SMG of C57BL/6 mice through activating the β-catenin signaling pathway, inhibiting cell apoptosis, and alleviating oxidative stress.

  • Abstract P2-10-01: Real-world Experience with CDK4/6 Inhibitors in Hormone Receptor-Positive Metastatic and Recurrent Breast Cancer in Asian Population

    Clinical Cancer Research · 2025-06-13

    article

    Abstract Purpose: Cyclin-dependent kinase 4/6 inhibitors (CDK4/6i) combined with endocrine therapy have demonstrated significant clinical benefits in progression-free and overall survival. This study investigates the outcomes associated with two kinds of CDK4/6i, palbociclib and ribociclib, in patients with hormone receptor (HR) positive metastatic and relapsed breast cancer to inform real-world evidence of treatment strategies. Methods: This retrospective study included 340 Taiwanese patients with HR-positive advanced breast cancer from the Taipei Veterans General Hospital, between 2018 and 2023. We analyzed patient characteristics, treatment strategies and outcomes associated with two CDK4/6i. The efficacy of patients who experienced economic burden and interrupted CDK4/6i treatment after 2 years of national health insurance(NHI) reimbursement was also investigated. Results: Patients receiving ribociclib and palbociclib showed no significant differences in age, histology, body mass index(BMI), or pathologic status. The distributions of disease status and endocrine therapy partners were comparable between the two groups. The dose reduction rate was similar, while patients with palbociclib tended to discontinue CDK4/6i usage, and those with ribociclib tended to switch to the other CDK4/6i or endocrine partners. Patients with palbociclib had a higher prevalence of prior chemotherapy for advanced diseases. There was no significant difference in progression-free survival (PFS) between the two CDK4/6i in the first-line setting. Adverse prognostic factors were increasing HER2 IHC score, higher Ki-67 levels, visceral and liver metastasis, prior chemotherapy, and endocrine therapy resistance while higher BMI, bone-only metastasis, and letrozole treatment were associated with a lower risk of progression. This limited follow-up time in our study was insufficient to assess the outcomes of patients treated with interrupted CDK4/6i for up to two years under the NHI reimbursement policy. Conclusion: Treatment outcomes between the two types of CDK4/6i did not differ significantly, indicating the safety and efficacy of CDK4/6i for the Asian population. Ribociclib and palbociclib showed similar efficacy in PFS in the real-world setting. Citation Format: Bo-Fang Chen, Chi-Cheng Huang, Ling-Ming Tseng, Yi-Fang Tsai, Ta-Chung Chao, Pei-Ju Lien, Yen-Shu Lin, Chin-Jung Feng, Yen-Jen Chen, Han-Fang Cheng, Chun-Yu Liu, Jiun-I Lai. Real-world Experience with CDK4/6 Inhibitors in Hormone Receptor-Positive Metastatic and Recurrent Breast Cancer in Asian Population [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr P2-10-01.

  • Beyond Swipe and Tap: Cultivating Mobile Accessibility Awareness and Knowledge in Computing Disciplines

    2025-08-21 · 1 citations

    article
  • ITGB2 Gene Gain as a Prognostic Marker for Early Recurrence in Luminal HER2-Negative High Proliferative Breast Cancer

    Cancer Research and Treatment · 2025-12-04

    articleOpen access1st authorCorresponding

    Purpose: Early recurrence reflects aggressive disease and poor prognosis, with patterns varying across breast cancer subtypes. Early recurrence is a major challenge in luminal HER2-negative high proliferative breast cancer, yet reliable biomarkers remain limited. Identifying biomarkers associated with early recurrence is therefore of significant interest. Materials and Methods: Forty-two patients with stage I-III luminal HER2-negative, high-proliferative breast cancer were enrolled in the VGHTPE cohort and underwent targeted next-generation sequencing using the ACTOnco Comprehensive Cancer Panel. Findings were validated in The Cancer Genome Atlas (TCGA) and METABRIC datasets, and in vitro experiments were conducted to assess cell aggressiveness. Results: We examined 42 patients, including 15 who experienced recurrence within five years and 27 who did not. Next-generation sequencing revealed that the most prevalent alterations in this luminal HER2-negative, high-proliferative breast cancer cohort were mutations in PIK3CA and TP53, and amplifications/gains in MCL1, MYC, and KLF6. Patients with recurrence within five years exhibited a higher frequency of ITGB2 gain compared with non-recurrent patients. Validation in the TCGA and METABRIC cohorts confirmed the prognostic significance of ITGB2 gain in ER+/HER2- breast cancers. Gene Set Enrichment Analysis indicated that high ITGB2 expression was enriched in KRAS and EMT pathways. In vitro experiments further demonstrated that ITGB2 knockdown led to inactivation of Akt and ERK signaling and suppression of cell aggressiveness in ER+/HER2- breast cancer cells. Conclusion: ITGB2 gain represents an unfavorable prognostic marker for early recurrence in luminal HER2-negative high proliferative breast cancer, with potential therapeutic implications.

  • Developing flow cytometry crossmatch thresholds: a statistical evaluation

    Human Immunology · 2025-09-01

    articleSenior author
  • Association of Epigenetic Age Acceleration and Mitochondrial DNA‐Based Aging Metrics Provides Insights Into Mechanisms of Aging‐Related Diseases

    UNC Libraries · 2025-12-12

    articleOpen access

    Investigating the interplay between mitochondrial DNA (mtDNA) variations and epigenetic aging metrics may elucidate biological mechanisms associated with age-related diseases. We estimated epigenetic age acceleration (EAA) metrics from DNA methylation data and derived mtDNA metrics, including heteroplasmic variants and mtDNA copy number (mtDNA CN) from whole genome sequencing. Linear regressions and meta-analyses were conducted to assess associations between EAA and mtDNA metrics, adjusting for chronological age, self-identified sex, and other covariates in 6,316 participants (58% female, 41% non-White Americans). Mediation analysis was conducted to examine whether EAA mediated the relationship between mtDNA CN and metabolic traits. A higher burden of rare heteroplasmic variants was associated with accelerations of first-generation EAA metrics, while a lower level mtDNA CN was associated with accelerations of second- and third-generation EAA metrics. For example, one standard deviation (SD) higher MSS, a score based on the predicted functions of rare heteroplasmic variants, was associated with a 0.22-year higher EAA by the Hannum method (p = 1.3E-6) among all participants, while one SD lower mtDNA CN was associated with higher DunedinPACE (β = -0.005, p = 6.0E-4). No significant association was observed between the heteroplasmy burden of common variants and EAAs. Furthermore, we observed DunedinPACE mediated 11.1% and 10.8% of the associations of mtDNA CN with obesity and T2DM in older FHS participants, respectively. Our analysis indicated that higher levels of heteroplasmy burden of rare variants and lower mtDNA CN were associated with accelerated epigenetic aging, and these associations showed stronger magnitudes among older participants.

  • 33. GENETIC REGULATION OF HUMAN BRAIN RNA EDITING IN EAST ASIAN

    European Neuropsychopharmacology · 2025-10-01

    articleOpen access

    RNA editing in the human brain exhibits a highly dynamic and complex regulatory landscape. With the rapid advancement of high-throughput sequencing technologies, many studies have focused on using edQTLs to explain GWAS signals. However, current studies are largely centered on European populations and primarily investigate genetic regulation at the level of individual editing sites. This approach overlooks the population-specific nature of RNA editing. Moreover, multiple studies have shown that RNA editing sites often occur in clusters and may be subject to coordinated regulation. To address those gaps, we aimed to construct the first large-scale RNA editing regulatory map in East Asians and proposed a novel concept of regional editing QTLs (redQTLs), treating regional average editing levels as an endophenotype for QTL analysis. We conducted whole-genome and transcriptome sequencing on 546 samples from the DLPFC of East Asian individuals. For cross-population comparison, we also included 429 matched DNA and RNA sequencing samples from the European BrainGVEx cohort. RNA editing sites were identified by integrating novel candidates from REDItools2 with known sites from public databases. We then performed edQTL mapping to assess the genetic regulation of RNA editing and compared the results with those from the European cohort. Co-localization analyses were conducted to fine-map SCZ GWAS signals. Finally, based on the average RNA editing level within each region, we performed regional RNA editing quantitative trait locus (redQTL) mapping to explore their coordinated genetic regulation, which reflects the biologically clustered nature of RNA editing. (1) We constructed the first large-scale map of genetic regulation of RNA editing in the East Asian population. 18,648 high-confidence RNA editing sites were identified in this study, and 504,500 edQTLs were identified by QTL analysis, of which 473,637 were East Asian–specific. (2) The edQTL repetition rate within the same population was higher than that between different populations(π1(EUR-EAS) = 0.73 and π1(EAS-EAS) = 0.86). Population-specific edQTL differences were mainly driven by allele frequency variation. Shared signals showed consistent regulatory directions. (3) Colocalization analysis identified 20 schizophrenia-associated editing sites in East Asians, including four novel candidate genes (MPHOSPH9, CSMD1, MFSD13A, KCNIP4), and the sites chr7_137067935 and chr7_137067938 were shared risk sites for both populations, with DGKI and UBE2D3 being the shared risk genes. (4) A total of 3,096 RNA editing regions and 136,286 redQTL were identified in the East Asian population by combining editing correlation, spatial proximity, and other factors. (5) redQTL showed a higher cross-population sharing rate compared to edQTL (from 12.8% to 39.7%). Notably, in the European population, we identified a novel SCZ-colocalized editing region (chr7_38764259_ld) that did not show significant signals in the edQTL analysis. This study constructed the first large-scale brain edQTL map in East Asian populations. It advances our understanding of population-specific mechanisms underlying RNA editing. Through colocalization analysis with neuropsychiatric disorders such as schizophrenia (SCZ), we identified four East Asian–specific SCZ-colocalized genes that have not been reported in previous studies. Additionally, our redQTL analysis further explores the coordinated genetic regulation of RNA editing.

  • Impact of common variants on brain gene expression from RNA to protein to schizophrenia risk

    Nature Communications · 2025-11-28

    articleOpen access

    Genetic variants influencing gene expression have been extensively studied at the transcriptional level. How these variants affect downstream processes remains unclear. We quantitated ribosome occupancy in prefrontal cortex samples from the BrainGVEX cohort and integrated these data with transcriptomic and proteomic profiles from the same individuals. Through cis-QTL mapping, we identified genetic variants associated with transcript level (eQTLs), ribosome occupancy (rQTLs), and protein level (pQTLs). Notably, only 34% of eQTLs have their effects propagated to the protein levels, suggesting widespread post-transcriptional attenuation. Using both a gene-based approach and a variant-based approach we identified omics-specific QTLs that associated with brain disorder GWAS signals and found the majority of them to be driven predominantly by transcriptional regulation. Consistently, using a TWAS approach, we identified 74 SCZ risk genes across the three omics layers, 52 were discovered using transcriptome with 68% showing limited impact on protein expression. Our findings indicated that many disease-associated variants act through regulatory mechanisms that do not lead to an observable impact on the protein level. By integrating a ribo-seq dataset of 195 individuals with RNA-seq and proteomics data from the same cohort, researchers found prevalent post-transcriptional attenuation of eQTL effects and identified omic-specific QTLs associated with schizophrenia.

  • Proteomics networks linking diet to cardiometabolic risk factors: the Framingham Heart Study

    American Journal of Clinical Nutrition · 2025-12-01 · 1 citations

    articleOpen access
  • EP.13.26 Gut Microbiota and Metabolites as Potential Biomarkers for Immunotherapy Efficacy in Patients With Extensive-Stage Small Cell Lung Cancer

    Journal of Thoracic Oncology · 2025-10-01

    article

Recent grants

Frequent coauthors

Labs

Education

  • Ph.D., Medical Genetics

    Central South University

    1998
  • M.S., Biology

    Xiamen University

    1994
  • B.S., Biology

    Wuhan University

    1991
  • Resume-aware match score
  • Save to shortlist
  • AI-drafted outreach

See your match with Chunyu Liu

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