Chin-Fu Hsiao

National Health Research Institutes, Miao-li-chieh, Taiwan, Taiwan

Are you Chin-Fu Hsiao?

Claim your profile

Publications (67)227.81 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: Methadone is a synthetic opioid that binds to the κ-opioid receptor with a low affinity. This study tested the hypotheses that the genetic polymorphisms in the κ-opioid receptor 1 (OPRK1) gene region are associated with methadone treatment responses in a Taiwan methadone maintenance treatment (MMT) cohort. Seventeen single nucleotide polymorphisms (SNPs) in OPRK1 were selected and genotyped on DNA of 366 MMT patients. Six SNPs from rs7843965 to rs1051660 (intron 2 to exon 2) were significantly associated with body weight (P < 0.007). A haplotype of 4 SNPs rs7832417-rs16918853-rs702764-rs7817710 (exon 4 to intron 3) was associated with bone or joint aches (P ≤ 0.004) and with the amount of alcohol use (standard drinks per day; global P < 0.0001). The haplotype rs10958350-rs7016778-rs12675595 was associated with gooseflesh skin (global P < 0.0001), yawning (global P = 0.0001), and restlessness (global P < 0.0001) withdrawal symptoms. The findings suggest that genetic polymorphisms in OPRK1 were associated with the body weight, alcohol use, and opioid withdrawal symptoms in MMT patients.
    Journal of clinical psychopharmacology 02/2014; · 5.09 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: In recent years, global collaboration has become a conventional strategy for new drug development. To accelerate the development process and shorten approval time, the design of multiregional clinical trials (MRCTs) incorporates subjects from many countries around the world under the same protocol. After showing the overall efficacy of a drug in all global regions, one can also simultaneously evaluate the possibility of applying the overall trial results to all regions and subsequently support drug registration in each of them. Several statistical methods have been proposed for the design and evaluation of MRCTs. Most of these approaches, however, assume a common variability of the primary endpoint across regions. In practice, this assumption may not be true, due to differences across regions (e.g., differences in ethnic factors and/or medical culture/practice). In this article, we use a random-effect model for modeling heterogeneous variability across regions for the design and evaluation of MRCTs. We also address consideration on the determination of the number of subjects in a specific region to establish the consistency of treatment effects between the specific region and the entire group.
    Journal of Biopharmaceutical Statistics 01/2014; 24(2):254-271. · 0.73 Impact Factor
  • Source
    Chieh Chiang, Chin-Fu Hsiao, Jen-Pei Liu
    [Show abstract] [Hide abstract]
    ABSTRACT: Statistical criterion for evaluation of individual bioequivalence (IBE) between generic and innovative products often involves a function of the second moments of normal distributions. Under replicated crossover designs, the aggregate criterion for IBE proposed by the guidance of the U.S. Food and Drug Administration (FDA) contains the squared mean difference, variance of subject-by-formulation interaction, and the difference in within-subject variances between the generic and innovative products. The upper confidence bound for the linearized form of the criterion derived by the modified large sample (MLS) method is proposed in the 2001 U.S. FDA guidance as a testing procedure for evaluation of IBE. Due to the complexity of the power function for the criterion based on the second moments, literature on sample size determination for the inference of IBE is scarce. Under the two-sequence and four-period crossover design, we derive the asymptotic distribution of the upper confidence bound of the linearized criterion. Hence the asymptotic power can be derived for sample size determination for evaluation of IBE. Results of numerical studies are reported. Discussion of sample size determination for evaluation of IBE based on the aggregate criterion of the second moments in practical applications is provided.
    PLoS ONE 01/2014; 9(10):e109746. · 3.53 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Abstract Methadone maintenance therapy is an established treatment for heroin dependence. This study tested the influence of functional genetic polymorphisms in CYP2C19 gene encoding a CYP450 enzyme that contributes to methadone metabolism on treatment dose, plasma concentration, and side effects of methadone. Two single nucleotide polymorphisms (SNPs), rs4986893 (exon 4) and rs4244285 (exon 5), were selected and genotyped in 366 patients receiving methadone maintenance therapy in Taiwan. The steady-state plasma concentrations of both methadone and its EDDP metabolite enantiomers were measured. SNP rs4244285 allele was significantly associated with the corrected QT interval (QTc) change in the electrocardiogram (p=0.021), and the Treatment Emergent Symptom Scale (TESS) total score (p=0.021) in patients who continued using heroin, as demonstrated with a positive urine opiate test. Using the gene dose (GD) models where the CYP2C19 SNPs were clustered into poor (0 GD) versus intermediate (1 GD) and extensive (2 GD) metabolizers, we found that the extensive metabolizers required a higher dose of methadone (p=0.035), and showed a lower plasma R-methadone/methadone dose ratio (p=0.007) in urine opiate test negative patients, as well as a greater QTc change (p=0.008) and higher total scores of TESS (p=0.018) in urine opiate test positive patients, than poor metabolizers. These results in a large study sample from Taiwan suggest that the gene dose of CYP2C19 may potentially serve as an indicator for the plasma R-methadone/methadone dose ratio and cardiac side effect in patients receiving methadone maintenance therapy. Further studies of pharmacogenetic variation in methadone pharmacokinetics and pharmacodynamics are warranted in different world populations.
    Omics: a journal of integrative biology 09/2013; · 2.29 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: This is a single-blind, parallel, flexible-dose study to compare the efficacy and tolerability of escitalopram and paroxetine in the treatment of patients with major depressive disorder. We recruited 399 patients from the outpatient clinics of five hospitals in northern Taiwan. Patients were administered either escitalopram (10-30 mg) or paroxetine (20-40 mg) according to the judgment of clinicians. These patients were assessed using the Hamilton Rating Scale for Depression (HAM-D) and the Hamilton Rating Scale for Anxiety at weeks 0, 1, 2, 4, 6, and 8. A total of 302 patients fulfilled the evaluation criteria and were included in a statistical analysis. We found that escitalopram induced more significant symptom reduction and response rate in terms of the mean HAM-D scores at week 6 (P<0.05) and week 8 (P<0.05) than paroxetine, but that there were no significant differences between the two groups in the remission rate. Escitalopram induced significantly less frequency of adverse effects of weakness (P<0.01), nausea and vomiting (P<0.001), drowsiness (P<0.01) as well as somnolence (P<0.01) than paroxetine, although all these side effects were mild and tolerable. However for a more definitive result, future prospective trials with the inclusion of a placebo group and a double-blind design are needed. In patients who did not have severe depression (HAM-D score at baseline<21), but not in severely depressed patients, escitalopram was statistically superior to paroxetine, as shown by the mean change in the HAM-D score.
    International clinical psychopharmacology 07/2013; · 3.35 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Aim: The liver CYP1A2 enzyme may metabolize antidepressant escitalopram (S-CIT) to S-desmethylcitalopram (S-DCIT) and S-didesmethylcitalopram (S-DDCIT). This study tested whether genetic polymorphisms in the CYP1A2 gene are associated with the treatment responses to S-CIT. Materials & methods: Ten SNPs in CYP1A2 were selected and genotyped in 158 patients under S-CIT treatment. The serum levels of S-CIT and its metabolites were measured by HPLC. Results:CYP1A2 SNPs rs2069521, rs2069526, rs4646425 and rs4646427 are significantly associated with the metabolic ratios of S-DDCIT/S-DCIT (p = 0.002, 0.018, 0.008 and 0.004, respectively) at week 2 of treatment. Carriers of the allele types associated with higher S-DDCIT/S-DCIT ratios had more severe side effects. Conclusion: These results suggest that genetic variants in CYP1A2 may be indicators for S-CIT metabolism and that the fast metabolizers may experience more severe adverse reactions in the early stages of S-CIT treatment. Original submitted 27 December 2012; Revision submitted 15 May 2013.
    Pharmacogenomics 07/2013; 14(10):1191-201. · 3.86 Impact Factor
  • Journal of clinical psychopharmacology 02/2013; 33(1):137-140. · 5.09 Impact Factor
  • Yufen Huang, Wan-Jung Chang, Chin-Fu Hsiao
    [Show abstract] [Hide abstract]
    ABSTRACT: To accelerate the drug development process and shorten approval time, the design of multiregional clinical trials (MRCTs) incorporates subjects from many countries/regions around the world under the same protocol. After showing the overall efficacy of a drug in all global regions, one can also simultaneously evaluate the possibility of applying the overall trial results to all regions and subsequently support drug registration in each of them. In this paper, we focus on a specific region and establish a statistical criterion to assess the consistency between the specific region and overall results in an MRCT. More specifically, we treat each region in an MRCT as an independent clinical trial, and each perhaps has different treatment effect. We then construct the empirical prior information for the treatment effect for the specific region on the basis of all of the observed data from other regions. We will conclude similarity between the specific region and all regions if the posterior probability of deriving a positive treatment effect in the specific region is large, say 80%. Numerical examples illustrate applications of the proposed approach in different scenarios. Copyright © 2013 John Wiley & Sons, Ltd.
    Pharmaceutical Statistics 01/2013; · 0.99 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Co-heritability of hypertension and insulin resistance (IR) within families not only implies genetic susceptibility may be responsible for these complex traits but also suggests a rational that biological candidate genes for hypertension may serve as markers for features of the metabolic syndrome (MetS). Thus we determined whether the T323C polymorphism (rs5333) of endothelin type A (ETA) receptor, a predominant receptor evoking potent vasoconstrictive action of endothelin-1, contributes to susceptibility to IR-associated hypertension in 1694 subjects of Chinese and Japanese origins. Blood pressures (BPs) and biochemistries were measured. Fasting insulin level, insulin-resistance homeostasis model assessment (HOMAIR) score, and area under curve of insulin concentration (AUCINS) were selected for assessing insulin sensitivity. Genotypes were obtained by methods of polymerase chain reaction-restriction fragment length polymorphism. Foremost findings were that minor allele frequency of the T323C polymorphism was noticeable lower in our overall Asian subjects compared to multi-national population reported in gene database; moreover both the genotypic and allelic frequencies of the polymorphism were significantly different between the two ethnic groups we studied. The genotype distributions at TT/TC/CC were 65, 31, 4% in Chinese and 51, 41, 8% in Japanese, respectively (p < 0.0001). Additionally, carriers of the C homozygote revealed characteristics of IR, namely significantly higher levels of fasting insulin, HOMAIR score, and AUCINS at 29.3, 35.3, and 39.3%, respectively, when compared to their counterparts with TT/TC genotypes in Chinese. Meanwhile, the CC genotype was associated with a higher level of high density lipoprotein cholesterol in Japanese. No association of the polymorphism with BP was observed. This study demonstrated for the first time that T323C polymorphism of ETA receptor gene was associated with an adverse insulin response in Chinese and a favorite atherogenic index in Japanese.
    Frontiers in Endocrinology 01/2013; 4:172.
  • Journal of clinical psychopharmacology 12/2012; · 5.09 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Hormone replacement therapy (HRT) and EGFR single nucleotide polymorphisms (SNPs) have been reported as risk factors for lung cancer in never smokers. We investigate the interaction of EGFR SNPs and HRT for lung adenocarcinoma risk in never-smoking women. This study included 532 never-smoking female lung adenocarcinoma patients and 532 controls, with EGFR SNPs retrieved from a genome-wide association study. The associations of EGFR SNPs with the lung adenocarcinoma risk were estimated by multivariate-adjusted logistic regression. The Haploview program was used to select tagged EGFR SNPs interacted with HRT and construct haplotype blocks. The Benjamini and Hochberg method was used to reduce the multiple testing effects. Among 84 EGFR SNPs retrieved, 11 tagging EGFR SNPs showed an interaction with HRT and lung adenocarcinoma risk, which were mostly located near the tyrosine kinase domain. Eight of the tagged SNPs were in two haplotype blocks. The interactions between HRT and numbers of protective EGFR SNP genotypes are significant in both blocks (p for interaction = 0.0004 and 0.0032, respectively). A trend of decrease in lung adenocarcinoma risk was found in subjects with HRT harboring an increasing number of protective EGFR SNP genotypes in both blocks (p = 0.0032 and 0.0046, respectively). In conclusion, HRT use may modify the association of EGFR SNPs with lung adenocarcinoma risk. The EGFR SNPs have a cumulative effect on decreasing lung adenocarcinoma risk in never-smoking women with HRT use.
    Carcinogenesis 12/2012; · 5.64 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Majority of the heroin-dependent patients smoke cigarettes. Although it has been reported that the OPRM1 genetic polymorphism is associated with the brain mu-opioid receptor binding potential in cigarette smokers, there is no direct evidence showing the impact of plasma cotinine, a nicotine metabolite, on treatment responses to methadone maintenance treatment (MMT). In this study, we aimed to test the hypothesis that the genetic polymorphisms in the OPRM1 are associated with the methadone treatment responses and the severity of cigarette smoking directly measured by the plasma concentration of cotinine in a Taiwanese MMT cohort. Fifteen OPRM1 single-nucleotide polymorphisms (SNPs) were selected and genotyped on DNA samples of 366 MMT patients. Plasma concentrations of cotinine were measured by cotinine enzyme-linked immunosorbent assay. The plasma cotinine concentration had positive correlation with concentrations of methadone (P=0.042) and its metabolite 2-ethylidene-1,5-dimethyl-3,3-diphenyl-pyrrolidine (P=0.037). Methadone treatment non-responders, defined by a positive urine morphine test, had a higher plasma concentration of cotinine (P=0.005), but a lower plasma concentration-to-dose ratio of both R- and S-methadone (P=0.001 and 0.012, respectively) than the responders. OPRM1 genetic variants, rs1074287, rs6912029, rs1799971, rs12209447, rs510769, rs3798676, rs553202, rs7748401, rs495491, rs10457090, rs589046, rs3778152 and rs563649, were significantly associated with the plasma concentration of cotinine when using recessive model for genotypes (general linear model (GLM), P<0.038; false discovery rate (FDR)<0.035) and additive model for allele types (GLM, P<0.03; FDR<0.049) in association analyses. The G allele carriers of SNP rs1799971 (A118G) on exon 1 of OPRM1 gene had a lower plasma cotinine concentration than the A allele carriers (GLM, P=0.029). OPRM1 genetic polymorphisms are associated with the plasma concentration of cotinine in a Taiwanese MMT cohort. Carriers with the major allele of SNP rs1799971 had a higher plasma cotinine concentration.Journal of Human Genetics advance online publication, 6 December 2012; doi:10.1038/jhg.2012.139.
    Journal of Human Genetics 12/2012; · 2.53 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: To identify common genetic variants that contribute to lung cancer susceptibility, we conducted a multistage genome-wide association study of lung cancer in Asian women who never smoked. We scanned 5,510 never-smoking female lung cancer cases and 4,544 controls drawn from 14 studies from mainland China, South Korea, Japan, Singapore, Taiwan and Hong Kong. We genotyped the most promising variants (associated at P < 5 × 10(-6)) in an additional 1,099 cases and 2,913 controls. We identified three new susceptibility loci at 10q25.2 (rs7086803, P = 3.54 × 10(-18)), 6q22.2 (rs9387478, P = 4.14 × 10(-10)) and 6p21.32 (rs2395185, P = 9.51 × 10(-9)). We also confirmed associations reported for loci at 5p15.33 and 3q28 and a recently reported finding at 17q24.3. We observed no evidence of association for lung cancer at 15q25 in never-smoking women in Asia, providing strong evidence that this locus is not associated with lung cancer independent of smoking.
    Nature Genetics 11/2012; · 35.21 Impact Factor
  • Source
    Chi-Tian Chen, H M James Hung, Chin-Fu Hsiao
    [Show abstract] [Hide abstract]
    ABSTRACT: To speed up drug development to allow faster access to medicines for patients globally, conducting multiregional trials incorporating subjects from many countries around the world under the same protocol may be desired. Several statistical methods have been proposed for the design and evaluation of multiregional trials. However, in most of the recent approaches for sample size determination in multiregional trials, a common treatment effect of the primary endpoint across regions is usually assumed. In practice, it might be expected that there is a difference in treatment effect due to regional difference (e.g., ethnic difference). In this article, a random effect model for heterogeneous treatment effect across regions is proposed for the design and evaluation of multiregional trials. We also address consideration of the determination of the number of subjects in a specific region to establish the consistency of treatment effects between the specific region and the entire group.
    Journal of Biopharmaceutical Statistics 09/2012; 22(5):1037-50. · 0.73 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: One of the challenges of multiregional drug development program is to design and analyze a multiple regional clinical trial with the objective being to satisfy different regional requirements on primary endpoints. Considered in this article is a multiregional clinical trial (MRCT) designed to test for two primary endpoints. Data of a regular fixed-size well-controlled parallel arm trial are used to test for two null hypotheses in terms of two distinct yet correlated endpoints. The two hypotheses may be tested sequentially or simultaneously. Depending on the structure of the hypotheses to be tested and the understanding of type I error rate control, various scenarios of type I error rate adjustments may be applied. Furthermore, for the objective of getting approval from regional authorities for different primary endpoints, various sample size and power determinations may be applied. In this article, comparisons of different approaches are discussed systematically.
    Journal of Biopharmaceutical Statistics 09/2012; 22(5):1051-9. · 0.73 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: In 1998, the International Conference on Harmonization (ICH) published a guidance to facilitate the registration of medicines among ICH regions including the European Union, the United States, and Japan by recommending a framework for evaluating the impact of ethnic factors on a medicine's effect, such as its efficacy and safety at a particular dosage and dose regimen (ICH E5, 1998). The purpose of ICH E5 is not only to evaluate the ethnic factor influence on safety, efficacy, dosage, and dose regimen, but also more importantly to minimize duplication of clinical data and allow extrapolation of foreign clinical data to a new region. In this article, statistical methods for evaluation of bridging studies based on the concepts of consistency (Shih, 2001 ), reproducibility/generalizability (Shao and Chow, 2002 ), the weighted Z-tests for the design of bridging studies (Lan et al., 2005 ), and similarity between the new and original region based in terms of positive treatment effect (Hsiao et al., 2007 ) are studied. The relative merits and disadvantages of these methods are compared by several examples.
    Journal of Biopharmaceutical Statistics 09/2012; 22(5):903-15. · 0.73 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To test whether the genetic polymorphisms within the gene encoding the UGT2B7 gene may have an impact on methadone treatment. Twelve SNPs in UGT2B7 were selected. 366 methadone maintenance treatment patients in Taiwan were recruited and genotyped. In a genotype recessive model, rs6600879, rs6600880, rs4554144, rs11940316, rs7438135, rs7662029, rs7668258, rs7439366, rs4292394 and rs6600893 showed significant associations with severity of withdrawal symptoms (permutation p < 0.002), pupil size (permutation p < 0.048) and tremor (permutation p < 0.008). Haplotypes of GATCAGCCGC and CTCTGATTCT were significantly associated with pupil size score and tremor score (p < 0.034). These results suggest that SNPs of the UGT2B7 gene may play important roles in opiate withdrawal symptoms.
    Pharmacogenomics 06/2012; 13(8):879-88. · 3.86 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: OBJECTIVE: Lung cancers that occur in never smokers differ from those that occur in smokers. We performed an analysis of potential epidemiological risk factors for lung cancer among never smokers. METHODS: In this hospital-based matched case-control study, all 1,540 matched case-control pairs were Han Chinese in Taiwan. The data on demographic characteristics, smoking habit, exposure to environmental tobacco smoke, medical history of lung diseases, family history of lung cancer, and female characteristics were collected from a structured questionnaire. A multiple conditional logistic regression was used to estimate odds ratios and 95 % confidence intervals after adjusting for possible confounders. RESULTS: Overall, several epidemiological factors of lung cancer in never smokers were different between males and females. For the female population, subjects who were exposed to environmental tobacco smoke (OR = 1.39, 95 % CI = 1.17-1.67) with a history of pulmonary tuberculosis and with family history of lung cancer in first-degree relatives (OR = 2.44, 95 % CI = 1.79-3.32) had higher risk of lung cancer, while subjects with a history of hormone replacement therapy and using fume extractors for those who cooked were protective. For the male population, only subjects with family history of lung cancer in first-degree relatives (OR = 2.77, 95 % CI = 1.53-5.01) were significantly associated with risk of lung cancer. CONCLUSION: This study provides insights about the epidemiological factors of lung cancer in never smokers, adding to existing evidence that family history of lung cancer and environmental tobacco smoke may moderate lung cancer risk.
    Cancer Causes and Control 05/2012; · 3.20 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: In recent years, global collaboration has become a conventional strategy for new drug development. To accelerate the development process and shorten approval time, the design of multi-regional clinical trials (MRCTs) incorporates subjects from many countries/regions around the world under the same protocol. After showing the overall efficacy of a drug in a global trial, one can also simultaneously evaluate the possibility of applying the overall trial results to all regions and subsequently support drug registration in each region. However, most of the recent approaches developed for the design and evaluation of MRCTs focus on establishing criteria to examine whether the overall results from the MRCT can be applied to a specific region. In this paper, we use the consistency criterion of Method 1 from the Japanese Ministry of Health, Labour and Welfare (MHLW) guidance to assess whether the overall results from the MRCT can be applied to all regions. Sample size determination for the MRCT is also provided to take all the consistency criteria from each individual region into account. Numerical examples are given to illustrate applications of the proposed approach.
    Pharmaceutical Statistics 04/2012; 11(4):295-9. · 0.99 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: INTRODUCTION: The hereditability of insulin resistance has been demonstrated in both familial and twin studies. The effects of renin-angiotensin-aldosterone system gene polymorphisms on insulin resistance remain inconclusive. METHODS: This is a sibling-based association study. Polymorphisms of renin-angiotensin-aldosterone system genes were examined in 1113 hypertension and 676 normotension siblings from Chinese and Japanese hypertensive families. The generalized estimation equations method was used to compare the differences in metabolic variables between hypertension and normotensive siblings. RESULTS: For the G-6A polymorphism of AGT, GG siblings had lower 2-h insulin than siblings carrying the A allele (p=0.006). Siblings with different variants of the angiotensin II type 1 receptor A1166C had no difference in metabolic variables. Siblings carrying the D allele of the angiotensin converting enzyme gene had higher levels of fasting glucose, fasting insulin, area under the curve of insulin levels and the homeostasis model assessment of insulin resistance than II siblings (all p<0.05). Lower levels of fasting glucose and 2-h glucose were observed in siblings with the T allele than their CC homozygotes for the C-344T polymorphism of CYP11B2 (p<0.05). Siblings carrying three high-risk genotypes of the angiotensin converting enzyme, angiotensinogen and CYP11B2 had higher fasting glucose level than siblings carrying no high-risk genotypes (p=0.011). CONCLUSION: Our comprehensive analysis of renin-angiotensin-aldosterone system gene polymorphisms demonstrates that the angiotensin converting enzyme and CYP11B2 gene polymorphisms are associated with insulin resistance in hypertensive families.
    Journal of Renin-Angiotensin-Aldosterone System 03/2012; · 2.29 Impact Factor

Publication Stats

612 Citations
227.81 Total Impact Points

Institutions

  • 2002–2014
    • National Health Research Institutes
      • Institute of Population Health Sciences
      Miao-li-chieh, Taiwan, Taiwan
  • 2013
    • National Chung Cheng University
      Chia-i-hsien, Taiwan, Taiwan
  • 2012
    • Duke University Medical Center
      • Department of Biostatistics and Bioinformatics
      Durham, NC, United States
    • National Taiwan University
      • College of Bio-Resources and Agriculture
      Taipei, Taipei, Taiwan
  • 2009–2012
    • National Institutes of Health
      • • Branch of Biostatistics and Bioinformatics (BB)
      • • Division of Cancer Epidemiology and Genetics
      Bethesda, MD, United States
    • Stanford University
      Palo Alto, California, United States
  • 2010
    • National Chiao Tung University
      • Institute of Statistics
      Hsinchu, Taiwan, Taiwan
  • 2008
    • Academia Sinica
      • Institute of Biomedical Sciences
      Taipei, Taipei, Taiwan
  • 2004
    • National Cheng Kung University
      • Department of Statistics
      Tainan, Taiwan, Taiwan