Determinants of personality and its association with risky health behaviors

양사라 서울대학교 2017 국내박사

Personality plays an important role in almost every aspects of individual’s psychosocial and physical health. Additionally, an individual’s personality affects health by predisposing the person to a higher chance of developing specific health behaviors. Attention to personality trait in clinical interventions is stressed in many reported health behavioral theories. Temperament and Character Inventory (TCI) is a comprehensive personality inventory that is widely used in behavioral genetics. The original theory suggested that temperament traits were under genetic influences, whereas character traits were gradually built by an interaction between temperaments and environment until early adulthood. Heritabilities of personality has been reported to be about 0.4 in average, but discrepancies in different ethnicities are present, and these estimates tend to be inflated when only twins are used in the study. Also, despite moderate heritability of 0.3-0.6 for most personality domains, even large powered studies with Caucasian samples were not successful in locating specific genetic variant that explains personality. Besides identifying the determinants of personality, confirming its relationship to risky health behaviors, specifically with behaviors that are theorized to have psychosocial base, may give important public health implications. Additionally, personality may mediate the effects of gene on these health behaviors. This study attempted to evaluate TCI by examining the genetic and environmental contributions to personality with particular attention to spousal effects. Also, a genome-wide search for both TCI domains and multivariable TCI-Five-Factor Model (FFM, for sub-samples) were conducted in a population cohort of Korea, where cultural environments are different from Western populations. Lastly, individual personality traits or their linear combinations were tested for associations with health behaviors, such as eating behaviors, nicotine dependency, and alcohol dependency. Interacting effects between personality and genetic variants, which are reported to be related to psychosocial traits, on these health behaviors were explored. This study includes a total of 3,479 individuals (1419 men, 690 families, 552 monozygotic twins, 119 dizygotic twins) from the Healthy twin study of Korea with detail epidemiologic, clinical information and TCI measures. Intraclass correlation coefficients (ICCs) and heritability were calculated to examine the genetic and shared environmental contributions to personality. Among these participants, 3,428 subjects with TCI measures were included in the univariate genome-wide association test. A total of 1,169 individuals (476 men) also fulfilled the FFM measures. Two platforms (Affymetrix Genome-Wide Human array 6.0, Illumina Infinium Humancore Beadchips) were used for genotyping and the markers were imputed using 1kG Asians. For univariate analysis, a family-based association test using mixed-effect variance component approach was conducted. For multivariable analysis of TCI-FFM model, multiple family-based quasi-likelihood score test (MFQLS) was performed to estimate the association of genetic variant to multiple phenotypes in linear mixed model. Linear association was tested for TCI traits and health behaviors adjusting for age, gender, education, income level, and familial correlation. For health behavior traits, Dutch Eating Behavior Questionnaire (DEBQ, n=3,444), Fagerstorm Test for Nicotine Dependence (FTND, n=1,192), and Alcohol Use Disorders Identification Test (AUDIT, n=2,431) were measured. Factor Analysis on particular personality traits (novelty seeking (NS), self-directedness (SD), Cooperativeness (CO)) that were commonly observed to be associated with the targeted health behaviors were conducted using principal axis analysis, and the identified personality patterns were also tested for associations with health behaviors. A total of 183 genetic variants that were reported to be related with neuro-psychological traits were tested to see if they have gene-by-personality (GxP) effect on the health behaviors. Moderate genetic contributions (0.15-0.44) were found for all TCI traits along with the evidence of shared environment (0.12-0.29) for harm avoidance (HA) and all characters. The ICCs of TCI in MZ pairs ranged 0.36-0.45. Spouses’ had little resemblance for temperament, whereas for character dimensions, spouses (0.27-0.38) were more similar than first degree relatives (0.10-0.27). Resemblance between spouses increased with duration of marriage for most characters and HA. When the growing similarities between spouses were compared with their MZ cotwins’ for subgroup of 84 trios, self-directedness (SD) of character showed even more similarities toward their spouses than cotwins as partnership duration increased (r=0.29). Univariate TCI domain analysis CO domain were associated with ADAMTS17 gene (Chromosome 15, p= 9.0e-8), but other domains did not reach genome-wide significance level. For agreeableness of FFM, CDH13 gene (Chromosome 16, p=2.8e-7) had marginal p-value, and the same variant had association with agreeableness in previous reported study (p=0.046). In multivariate analysis, SD (TCI) and neuroticism (FFM) showed significant genetic associations in RP11-274B18.4 gene (r=-0.56, Chromosome 9, p= 7.52e-9). All seven TCI traits showed association with some targeted behaviors. Among them, combination of high NS, low SD, and low CO was observed in emotional eating, external eating, nicotine dependence, and alcohol dependence. The subscales of these traits were factorized to obtain specific personality profile named Vulnerability, and its association with, as well as its’ ability to predict risky health behaviors was also observed. SNAP25 gene interacted with self-transcendence in its association to emotional eating, EVL gene interacted with SD on effect of external eating, and CACNA1C gene interacted with CO to affect FTND. The findings with regard to change in SD into late adulthood support the psychobiological theory of temperament and character, which suggests that both personality domains have distinct developmental trajectories despite equally large genetic influences. Also novel genetic loci were found to be associated personality traits, particularly in character dimensions. The findings from genome-wide search for associated gene illustrates that there may be different biological background in Koreans, and that a multiple measures of personality traits might better capture the genetic architecture of personality traits. Lastly, by finding the personality patterns and GxP effects that have shown increased risk in developing certain health behaviors, translating them into personalized medicine for effective prevention and intervention for risky health behaviors is possible.

Statistical Methods in Population Genetics and Viral Phylodynamics

Ki, Caleb ProQuest Dissertations & Theses University of Mich 2022 해외박사(DDOD)

Genetic sequences carry a wealth of information. Scientists and statisticians have utilized genetic variation data to answer a wide range of questions in evolutionary biology and epidemiology. With the advent of high throughput sequencing, the availability of genetic sequence data has exploded this century. While the unprecedented amount of genetic data available presents an opportunity to garner a deeper understanding about viruses and humans, making use of large volumes of genetic data is still a challenging problem.In what is to follow, we present three methods that tackle various problems analyzing genetic variation data. First, we introduce the framework known as the sequentially Markov coalescent (SMC), which enables likelihood based inference using hidden Markov models (HMMs) where the latent variables represent genealogies. While genealogies are continuous, HMMs are discrete, requiring SMC based methods to discretize genealogies. This discretization often leads to biased and noisy estimates of the population size history. We introduce a method that avoids the need for discretization leading to Bayesian and frequentist inference procedures that are faster and less biased than its predecessors.Additionally, while coalescent HMMs based on SMC can be decoded in linear time, there does not yet exist a linear time EM algorithm for coalescent HMMs based on SMC', the more accurate approximation. We present a linear time EM algorithm based on SMC. Advantages of this method include increased accuracy, computation time, uncertainty quantification, and ability to incorporate regularization.Lastly, we present a new approach for estimating transmission and recovery rates of viruses using genetic sequence data. With the outbreak of the SARS-CoV-2, there are millions of genomic sequences available to analyze, but few methods to exploit the information contained in these sequences. By integrating recent advances in Bayesian inference and differentiable programming with phylodynamics, we provide a method capable of estimating transmission, recovery, and sampling of pathogens using thousands of sequences. We apply our method to SARS-CoV-2 data and find that our estimates of the effective reproductive number closely match other estimates from methods based on public health data.

Statistical and Machine Learning Methods for the Analysis of Summary Statistics Derived From Large Genomic Datasets

Liao, Kevin Stanford University of Michigan ProQuest Dissertations & Th 2023 해외박사(DDOD)

Advancements in DNA sequencing over the past decade have transformed our ability to characterize genetic variation in large populations and study the genetics of many complex traits. For population geneticists, information on the genetic variation (i.e., which sites in the genome are mutated and at what frequency) alone is interesting as it allows for studying aspects of a population (e.g., demographic history, natural selection, and mutation rates). For statistical geneticists and genetic epidemiologists, the availability of phenotypic information in the same set of genetically sequenced individuals allow for studying the genetic basis of a complex trait. In this dissertation, I present three separate projects that leverage genetic information originating from DNA sequencing.In the first project I focused on analyzing genetic variation without consideration of a phenotype, as is often done in the field of population genetics to make inferences on demographic history or natural selection. A commonly used summary statistic of genetic variation for population genetics inference is the allele frequency spectrum. However, methods based on the allele frequency spectrum make a simplifying assumption: all sites are interchangeable (i.e., an A->T mutation is the same as a C->T) mutation. In this project, I first extended previous literature to show heterogeneity in the allele frequency spectrum exists across mutation types at finer levels of resolution. I then illustrated how inferences of demographic history and natural selection are impacted by the violation of this assumption.In the second project I focused on combining phenotypic information with genetic data through genome wide association studies (GWAS) and polygenic risk scores (PRS). GWAS estimate per-variant genetic effects on a complex trait, which can be used to summarize the genetic risk of that trait for an individual in PRS (constructed as the GWAS-weighted sum of their risk variants). However, PRS have a portability issue where phenotype predictions worsen as the ancestry of the target sample diverges from that of the GWAS sample. In admixed individuals, genome can be traced back to multiple ancestral populations and ancestry lies on a continuum. Such a continuum causes an ancestry dependence of PRS performance, as the PRS for samples whose ancestry better matches the external GWAS perform better. To help resolve this issue, I developed slaPRS, a stacking-based framework to integrate GWAS from multiple ancestral populations to construct polygenic risk scores (PRS) in admixed individuals. In simulations and real data, slaPRS performed well and reduced the ancestry dependence compared to existing approaches.In the third project I focused on how genetic-phenotypic associations are shared across two more phenotypes through pleiotropy. Pleiotropy can be characterized at resolutions including genome wide, regionally, or at the SNP/gene-level. One approach to studying pleiotropy is local genetic correlation (LGC), which quantifies the extent of genetic sharing in a local region through the similarity in GWAS effect sizes. However, one problem of LGC is that it remains unable to identify SNP or gene-level pleiotropy, making it impossible to identify which variants or genes in a region drive a signal of LGC. To resolve this issue, I developed LDSC-MIX, a Bayesian mixture of regression method to infer latent groups of likely shared causal variants across two traits. In simulations and real data, LDSC-MIX identified SNP sets recovering the true LGC and tested whether genes in a region are enriched for such SNPs.

Molecular genetic characterization of porcine epidemic diarrhea viruses (PEDV)

박성준 서울대학교 대학원 2012 국내박사

돼지 유행성 설사병 바이러스 (PEDV)는 막을 지닌 단일 가닥의 RNA 바이러스로서 코로나비리대 (Coronaviridae)에 속하며, 1978년 벨기에와 영국에서 최초 발생 보고 된 이후 많은 양돈 국가, 특히 유럽 및 아시아 국가에서 지속적으로 발생되고 있다. 국내에서는 1992년 PEDV가 처음 발견 되었으며, 그 후 PEDV 백신이 개발되어 흔히 사용되고 있다. 하지만 백신 정책 시행에도 불구하고 PEDV 감염으로 인한 양돈 농가의 피해는 여전히 심각하며, PEDV 감염을 보다 효과적으로 예방 및 통제하기 위해서는 PEDV의 특성뿐만 아니라 그것들의 국내 발생 현황에 대한 보다 정확한 이해가 필요하다. 따라서 본 연구에서는 현저히 다른 야생형 및 약독화형 PEDV를 비교 분석함으로써 PEDV의 분자 유전학적 특성을 규명하고, PEDV 국내 분리주에 대한 분자 역학, 유전적 다양성 및 다른 reference주와의 상관관계 분석을 통해 PEDV의 국내 발생 현황을 밝히고자 하였다. 돼지들에서 놀랄만한 병원성 차이를 보여준 이전 연구 결과를 바탕으로 attenuated DR13주의 전장의 S 유전자 서열을 parent DR13 및 reference주의 S 유전자 서열과 비교 분석하였다. 야생형 PEDV와 비교 분석시, attenuated DR13주의 S 유전자에서 50개의 염기변이가 발견되었으며, 이중 N-linked glycosylation site를 변화시키는 3개의 염기변이를 포함하여 총 20개의 염기변이가 결과적으로 attenuated DR13주의 S 단백질의 아미노산변이를 야기시키기 때문에 이 20개의 염기변이가 PEDV의 병원성에 있어서 굉장히 중요하다고 생각된다. PEDV의 유일한 액세서리 유전자인 ORF3 유전자 내에서 PEDV의 병원성 결정인자를 찾기 위해, parent DR13, attenuated DR13, KPED-9, P-5V 및 12개 필드 바이러스의 전장의 ORF3 유전자 서열을 분석하였다. CV777, Br1/87, LZC, parent DR13 및 12개 필드 바이러스를 포함하는 야생형 PEDV와 비교 분석시, 생독 백신주인 attenuated DR13, KPED-9, P-5V를 포함하는 약독화형 PEDV는 51개의 염기결손으로 인해 624개의 염기로 구성된 ORF3 유전자를 가졌으며, 이 유전자는 207개의 아미노산으로 구성된 23.4 kDa의 단백질을 암호화할 것으로 생각된다. 본 연구를 통해 ORF3 유전자 내에서의 염기결손이 최초 발견되었으며, 약독화형 PEDV의 ORF3 단백질의 아미노산변이가 이 염기결손으로 인해 야기된다는 점을 통해 이 염기결손은 PEDV의 병원성에 있어서 굉장히 중요할 것으로 생각된다. PEDV 국내 분리주의 유전적 다양성을 조사하고 국내에서 유행하는 PEDV를 밝히기 위해, 2002년부터 2005년까지 국내에서 분리된 45개 PEDV에 대해 항원결정부위를 포함하는 S 유전자 서열 분석을 실시하였다. S 유전자 분석 결과에 따르면, 국내 PEDV 분리주는 유전적 다양성을 나타냈으며, 3개의 소그룹 (G1-1, G1-2, G1-3)을 갖는 3개의 그룹 (G1, G2, G3)으로 나뉘었다. 분석에 사용된 45개의 PEDV 국내 분리주 가운데 37개의 국내 분리주가 CV777, Br1/87, JS-2004-2, KPED-9, P-5V, SM98-1, parent DR13 및 attenuated DR13주를 포함하는 G1에 속했으며, 특히 BI2357을 제외한 모든 G1 PEDV 국내 분리주가 중국주인 JS-2004-2를 포함하는 G1-1에 속했다. 각각의 PEDV 그룹은 그룹 특이적인 다수의 염기변이를 갖고 있었으며, 이와 같은 특성은 특정 그룹의 PEDV를 구별하는데 활용될 수 있다. 마지막으로 PEDV 국내 분리주의 분자 역학, 유전적 다양성 및 다른 reference주와의 상관관계 분석을 위해 2002년부터 2007년까지 국내에서 분리된 26개 PEDV의 전장의 M 및 ORF3 유전자 서열을 reference주의 M 및 ORF3 유전자 서열과 비교 분석하였다. M 및 ORF3 유전자 분석 결과에 따르면, 그룹 특이적인 염기변이를 갖고 있는 각각의 PEDV 그룹은, 이 특이적인 염기변이를 이용하여 구별될 수 있으며, 특히 ORF3 유전자 분석은 백신주 및 야외주 PEDV를 감별하는데 활용될 수 있다. M 및 ORF3 유전자에 대한 계통분석 결과에 따르면, 최근 국내에서 유행하는 PEDV는 중국주와 높은 상관관계를 나타낸 반면, 유럽주 및 국내에서 PEDV 예방을 위해 사용되고 있는 생독 백신주 (attenuated DR13, KPED-9, P-5V)와는 유전적으로 달랐다. 결론적으로, 본 연구를 통해 PEDV의 병원성과 관련 있을 것으로 생각되는 결정인자 (유전적 변이 및 결손)의 발견과 같은 PEDV의 분자 유전학적 특성이 규명되었으며, 규명된 특성은 특정 그룹의 PEDV 뿐만 아니라, 야외형 및 약독화형 PEDV의 감별을 위한 마커로써 활용될 수 있으며, 역유전학 (reverse genetics)을 이용한 PEDV의 병원성 결정인자 확인시, PEDV의 병원성과 관련된 결정인자의 범위를 좁히는데 있어서도 큰 도움을 줄 것이다. PEDV 국내 분리주는 유전적 다양성을 나타냈으며, 특히 최근 국내에서 유행하는 PEDV는 새로운 유전형으로써 PEDV 예방을 위해 국내에서 사용되고 있는 생독 백신주와 달랐다. 이 결과는 국내에서 PEDV 감염을 보다 효과적으로 예방하기 위해 새로운 형의 PEDV 백신이 필요할지 모를 거라는 문제를 제기한다. 본 연구는 PEDV의 특성 및 그것들의 국내 발생 현황에 대한 깊은 이해를 가능하게 하였으며, 저는 본 연구 결과가 국내 PEDV 감염을 보다 효과적으로 예방하고 통제하는데 있어 도움을 줄 수 있기를 기대합니다. Porcine epidemic diarrhea virus (PEDV), a member of the family Coronaviridae, is an enveloped, single-stranded RNA virus. PEDV was first reported in Belgium and the United Kingdom in 1978. Since the first identification of PEDV, outbreaks of PEDV infections have been reported in many swine-producing countries, notably in Europe and Asia. In Korea, the first identification of PEDV was reported in 1992. After the first identification of PEDV, PEDV vaccines were developed and have been commonly used. However, in spite of vaccination policy, PEDV infection has caused continuous damage and economic losses in pig farms. To effectively control and prevent PEDV infection, it is necessary to recognize more exactly the features of PEDVs and their current status in Korea. Therefore, in this study, molecular genetic characteristics of PEDVs were investigated by the comparison of the markedly different wild- and attenuated-type PEDVs, and current status of PEDVs in Korea was elucidated based on the molecular epidemiology and genetic diversity of Korean PEDV isolates, and their phylogenetic relationships to other PEDV reference strains. Based on previous studies shown the remarkable difference in pathogenicity in pigs, the full-length S gene sequence of attenuated DR13 strain was compared to that of other reference strains as well as the parent DR13. The S gene of attenuated DR13 had 50 nucleotide changes as compared to other wild-type PEDVs. Notably 20 nucleotide changes including 3 nucleotide changes, which cause alterations at the N-linked glycosylation sites, are thought to be significant for pathogenicity of PEDV in that they changed the amino acids in the S protein of the attenuated DR13. To reveal determinants related to pathogenicity of PEDV within ORF3 gene, the only accessory gene of PEDV, the full-length ORF3 gene sequences of parent DR13, attenuated DR13, KPED-9, P-5V, and 12 field isolates were analyzed. Attenuated-type PEDVs, including live vaccine strains (attenuated DR13, KPED-9, and P-5V), had respective ORF3 genes of 624 nucleotides encoding a protein of 207 amino acids with a predicted Mr of 23.4 kDa due to 51 nucleotide deletions as compared to those of wild-type PEDVs, including CV777, Br1/87, LZC, parent DR13, and 12 field isolates. Those deletions were first identified in this study and are suggested to be crucial for pathogenicity of PEDV because they lead to the amino acid changes in the ORF3 proteins of attenuated-type PEDVs. To investigate genetic diversity among Korean PEDV isolates and to find out more prevalent PEDVs in Korea, S gene sequences including epitope region of 45 PEDVs isolated in Korea from 2002 to 2005 were analyzed. The results of the S gene analysis showed that Korean PEDV isolates are genetically diverse, and are divided into three separate groups (G1, G2, G3), which have three subgroups (G1-1, G1-2, G1-3). Out of 45 Korean PEDV isolates, 37 Korean PEDV isolates were found to belong to the G1, including CV777, Br1/87, JS-2004-2, KPED-9, P-5V, SM98-1, parent DR13 and attenuated DR13. Notably, all G1 Korean isolates except BI2357 were closely related to the G1-1 including Chinese strain JS-2004-2. Each PEDV group had several unique nucleotide changes, and thus specific group can be differentiated from other groups by those nucleotide differences. Finally, full-length M and ORF3 gene sequences of 26 PEDVs isolated in Korea from 2002 to 2007 were analyzed to elucidate molecular epidemiology, genetic diversity, and phylogenetic relationships of Korean PEDV isolates to other reference strains. Genetic analysis of the M and ORF3 genes showed that each PEDV group had several unique nucleotide changes, and this indicated that specific group of PEDVs can be differentiated from other groups by those nucleotide differences. Especially, ORF3 gene analysis can be used for discrimination between vaccine- and wild-type PEDVs. Phylogenetic analysis of the M and ORF3 genes showed that recent prevalent Korean PEDV isolates have a close relationship to the Chinese field strains and differ genetically from the European PEDV strains and the live vaccine strains (attenuated DR13, KPED-9, and P-5V), which have been used for prevention of PEDV infection in Korea. In conclusion, molecular genetic characteristics of PEDVs, such as identification of determinants (genetic changes and deletions) suggested to be involved in pathogenicity of PEDV, were elucidated. These characteristics might be used as markers for differentiation of wild- and attenuated-type PEDVs as well as specific groups of PEDVs, and should help to localize determinants related to pathogenicity of PEDV as confirmed by using reverse genetics. The Korean PEDV field isolates were genetically diverse, and especially recent prevalent Korean PEDV field isolates represented a new genotype that differed from the genotype including the live vaccine strains, which have been used for prevention of PEDV infection in Korea. These results raise questions as to whether a new type of PEDV vaccine may be necessary for preventing PEDV infection more effectively in Korea. The present study allows a deep understanding on the features of PEDVs and their current status in Korea, and I expect that the results of this study will help to prevent and control infection of PEDV in Korea more effectively.

Genetics of tuberculosis resistance

Sobota, Rafal S Vanderbilt University 2015 해외박사(DDOD)

Mycobacterium tuberculosis (MTB) infection and subsequent tuberculosis (TB) is the second-leading cause of mortality from a single infectious agent worldwide, after the human immunodeficiency virus (HIV)1,2. In 2013, 9 million new cases of clinical tuberculosis were diagnosed and 1.5 million deaths were attributed to the disease2. An estimated 1.1 million new cases and 360,000 of the deaths occurred in people co-infected with HIV2. The immunosuppression resulting from HIV infection increases the risk of progression to active disease following new exposure to MTB, or reactivation of latent MTB in patients with prior infection 3,4. Sub-Saharan Africa is the location where most HIV/TB co-infection occurs, with 75% of all cases reporting coinfection1,2. The influence of host genetics on tuberculosis disease has been extensively studied, mostly in HIV-negative patients, and revealed that variation in pathways pertinent to macrophage and Type 1 helper T cell (TH1) signaling, among others, modulate disease risk5-11. Generally, HIV seropositive status has been viewed as a confounder in such studies and it is either adjusted for or used as an exclusion criterion. In the current project we present a novel hypothesis for studying resistance to either TB disease or MTB infection, using the immunosuppression of HIV-positive patients to identify an extreme phenotype. Namely, we posit that HIV-positive patients living in areas endemic for MTB who do not develop TB are resistant, and that this protection has a genetic component with an effect size large enough to permit using a smaller sample size than those seen in prior genome-wide association studies (GWAS) on TB10,11. The goal of the following chapters is to evaluate this hypothesis as it pertains to resistance to TB disease and MTB infection, in single variant and epistatic models. Background information pertinent to this project is described in part A of Chapter II, including clinical presentation, diagnostics, treatments, a summary of the worldwide burden of tuberculosis as well as a description of prior genetic variants associating with TB. A genome-wide association study of common genetic variants and with TB resistance in HIV-positive patients is described in Chapter III. Of note, I recruited patients on-site and isolated DNA in one of the cohorts, the DarDar vaccine trial extended follow up, in Dar es Salaam, Tanzania. I also isolated DNA from samples collected in two other studies, the DarDar Women's Nutrition Study from Tanzania and the Household Contacts Study from Kampala, Uganda. The samples for the Nutrition and Household Contacts Studies were made available to us by fellow investigators. In a study combining the cohorts, we found a common variant associating at the genome-wide significance threshold in the IL12B region. Linkage disequilibrium (LD) patterns in the region suggested that the region is conserved and integrated haplotype score analyses using sub-Saharan populations demonstrated that the LD block containing rs4921437 has undergone selection. The single nucleotide polymorphism (SNP) of interest is located in an area enriched for a histone acetylation mark often found in active regulatory elements, suggesting possible functionality and a genetic-epigenetic interaction at the site. Further studies of this interaction are warranted. Chapter IV describes two approaches used to evaluate the genetics of MTB infection. In Part A, we used a genome-wide approach to identify variants associating with MTB infection. In Part B, we validated and fine-mapped regions previously associated through genome-wide linkage analyses, SLC6A3 +/- 0.5 mb12, 2q1413, 2q21-2q24 133, 5p13-5q2213 and chromosome 11 p14.112. The immune response to MTB infection and resultant disease is complex 8,14-17. The molecular signaling profile and extent tissue involvement change over the course of disease. Therefore, it is likely that no single factor alone can adequately explain risk of TB disease or MTB infection risk. In Chapter V we present a study in which we examined multi-locus relationships between previously associated candidate genes and TB disease in Part A, and MTB infection in Part B, using Multifactor Dimensionality Reduction (MDR) software18,19. Chapter VII summarizes the conclusions of this project and proposes future directions. *Please refer to dissertation for footnotes. (Abstract shortened by UMI.).

Familial and hormonal factors affecting cancer risks in women

Lee, Jennifer Shuwen Stanford University 2008 해외박사(DDOD)

Context. Genetic susceptibility and hormonal exposures influence the development of many cancers, including cancers of the breast, ovary, and lymphatics. Chapters I and II utilize different study designs and methods to estimate population-based cancer risks related to genetic susceptibility and to reproductive and hormonal exposures in women. Little is known about the variation in the magnitudes of cancer risks in relatives according to the patient's BRCA1 or BRCA2 mutation status or other cancer-related characteristics. Better quantification of the magnitudes of familial cancer risks, as in Chapter I, may improve preventive genetic counseling and influence resultant decision-making among family members. Estrogen and other female hormone exposures may influence the risk of certain NHL subtypes by mediating immune responses. Challenges to epidemiological studies of this possibility include the need for sufficiently large, population-based studies of well-characterized diagnoses of NHL subtypes. This challenge is addressed in Chapter II. Chapter III focuses on a major methodological challenge in hormone and reproductive epidemiology: the determination of the best assay method to measure low serum estradiol levels as a biomarker of estrogen exposure in postmenopausal women. Methods. Chapter I utilizes a two-stage study design and Poisson regression models to estimate standardized incidence ratios for the magnitudes of breast and ovarian cancer risks in first-degree relatives of patients with breast or ovarian cancer. Chapter II utilizes detailed questionnaire information collected in a large population-based case-control study of confirmed incident NHL to test the hypothesis that exposure to estrogen and reproductive hormones provides protection against the development of NHL, overall and by major subtypes. Age- and multivariable-adjusted logistic regression models were conducted to estimate population-based odds ratios for NHL risk associated with reproductive and menstrual factors, oral contraceptive (OC) use, and menopausal hormone therapy (HT) use. In Chapter III, two major estradiol assay methods: indirect and direct immunoassays, were compared to mass spectrometry in an initial pilot study followed by a confirmatory study of postmenopausal women. Results and conclusions. In population-based studies of patients with breast or ovarian cancer, relatives of patients with and without BRCA1 mutations have breast and ovarian cancer risks that correlate with the patient's cancer site. In families of breast cancer patients without BRCA mutations, breast cancer risk depends on the patient's age at diagnosis. These patterns support the presence of genes and/or environmental factors that modify the risk specific to cancer site, in both carriers and noncarriers of BRCA1 and BRCA2 mutations. Future studies are warranted to identify the cancer site-specific modifying genes and environmental risk factors in both carriers and noncarriers of BRCA1 and BRCA2 mutations. In a population-based case-control study of NHL in women, greater parity and exogenous hormone therapy use were associated with a lower risk of the DLCL subtype whereas endogenous and exogenous reproductive hormone exposures were not associated with FL in women. These results suggest that endogenous and exogenous reproductive hormones confer different risks by NHL subtype and are associated with a reduced risk of DLCL in women. Future studies are warranted to confirm these results and to investigate interactions between female hormonal exposures and genetic susceptibility in the development of NHL subtypes in women. In a pilot study followed by a confirmatory study that compared indirect and direct assay methods to measure low serum estradiol levels, indirect radioimmunoassay with an initial extraction purification step was the preferred method to measure low estradiol levels typically seen in postmenopausal women. This result suggests that indirect methods should be used in epidemiological studies of estradiol as a biomarker of estrogen exposure in postmenopausal women. Future studies are warranted to confirm these results and to establish assay standardization.

Computational Methods for Large Scale Analysis of Microbial Genome Evolution and Application to Antibiotic Resistant Pathogens

Crawford, Ryan D ProQuest Dissertations & Theses University of Mich 2022 해외박사(DDOD)

Horizontal Gene Transfer is a powerful force shaping microbial evolution. This constant process by which genes are acquired into and excised from bacterial genomes enables an enormous capacity for rapid phenotypic evolution. HGT enables the dissemination of clinically important genes, including antibiotic resistance genes, genes mediating virulence, environmental persistence genes, and metabolic genes. Acquisition of these genes potentiates phenotypic evolution in several important contexts: increasing the capacity for transmission, enhancing the ability for infection, limiting the efficacy of antibiotic therapies, and facilitating the metabolism of new substrates. Methods to characterize the pathways by which these genes spread through bacterial populations are critical for understanding the evolution of these phenotypes and their implications for public health. In this dissertation, I develop a novel computational approach to generate core gene alignments for large numbers of bacterial genomes and implement two methods to characterize HGT events from bacterial whole-genome sequences. I then apply these methods to understand the dissemination of antimicrobial resistance genes and the evolution of carbohydrate utilization phenotypes in the microbiome. First, we developed cognac (Core Gene Alignment Concatenation), an open-source R package for generating concatenated, core gene alignments for microbial genomes. cognac rapidly identifies shared phylogenetic marker genes, creates gene alignments, and concatenates them into a single alignment for downstream phylogenetic analysis. We demonstrate that this method can efficiently handle extremely large whole-genome sequencing datasets of diverse bacterial lineages. Second, we sought to trace the spread of the KPC gene, a carbapenemase conferring broad-spectrum resistance to commonly used antibiotics for treating infections caused by Enterobacterales. Using comprehensive collections of clinical isolates from regional healthcare networks in three US states, we quantify the role of importation, clonal dissemination, and HGT on the total burden of KPC in these regions. To identify HGT events, we implemented a novel marker gene-based approach that enabled us to track KPC plasmid transfer using short-read data and identify HGT events occurring between circulating strains in the same region. Using this approach, we show that while the horizontal transfer of KPC frequently occurs in all three states, the strains and species involved and the overall contribution to the regional burden of KPC-carrying organisms differ substantially across the three states. Third, we investigated the role of HGT in common members of the human gut microbiome. We developed a novel method to identify ancestral HGT loci by identifying core genes with significantly greater than expected divergence from the assigned species and greater similarity to the putative donor species. We then characterized HGT loci with conserved synteny and collinearity between donor and recipient species that have enabled pan-genome expansion and evolution of new phenotypes. This approach illustrates that HGT is common between two closely related species of Bacteroides, with many loci exhibiting evidence of HGT. These data, in conjunction with molecular data, provide insight into the breadth and complexity of metabolism in the microbiome and the underlying genomic events that enable the evolution of complex phenotypes. In summary, this body of work establishes computational tools with broad application in computational genomics and genomic epidemiology: enabling phylogenetic analysis of large genomic datasets, identifying recent plasmid-mediated transfer occurring within and across regional healthcare networks, and identification of ancestral HGT loci carried on the chromosome mediating the development of complex phenotypes in the microbiome.

Pathogenic Variant Status of High-Risk Genes, Polygenic, Epidemiological Risk Factors, and Utilization of Risk Management Options in Breast Cancer

Guo, Boya University of Washington ProQuest Dissertations & 2023 해외박사(DDOD)

Breast cancer is a heterogeneous disease with multiple established risk factors, which include high-penetrance germline variants in cancer predisposition genes such as ATM, BRCA1, BRCA2, CHEK2, and PALB2. Additionally, individual and behavioral factors such as age at menarche, parity, number of births, age at first full-term pregnancy, breastfeeding, age at natural menopause, height, pre- and post-menopausal body mass index (BMI), use of menopausal hormone treatment (PMH) and oral contraceptives (OC), history of benign breast disease (BBD), smoking and alcohol consumption have been consistently observed to be associated with breast cancer risk and may potentially modify the risk associated with pathogenic variants (PV). However, existing gene-environment interaction (GE) studies of rare PVs in breast cancer predisposition genes have been limited by sample size and the number of interactions assessed. GE studies have rarely been replicated, due to issues related to statistical power, exposure measurement errors, and difficulties in harmonizing data across different studies. Further, breast cancer is often treated as one disease, without consideration of molecular subtypes, primarily defined by estrogen receptor (ER) status, ignoring well-known differences in etiologies. Further, women who have been identified to have an elevated lifetime risk of breast cancer through genetic testing for hereditary cancer syndromes can benefit from risk management options such as enhanced screening and preventive surgery. However, these recommended risk management strategies are underutilized in current clinical practice, and healthcare utilization patterns following genetic testing remain poorly understood, particularly in underserved populations facing access barriers to genetic services.In this dissertation, we explored how genetic, epidemiological, and behavioral factors collectively influence breast cancer risk. Specifically, we aimed to 1) examine GE interactions between high penetrance breast cancer susceptibility genes and well-established epidemiological risk factors; 2) assess the impact of modifiable epidemiological risk factors on breast cancer risk across categories defined by nonmodifiable risk factors including a polygenic risk score (PRS) of common genetic variants; 3) examine the uptake of recommended risk management strategies for breast cancer subsequent to the disclosure of genetic testing results.To achieve these aims, we first explored GE interactions on both multiplicative and additive scales in a sample of women drawn from the Cancer Risk Estimates Related to Susceptibility (CARRIERS) Consortium and the UK Biobank, comprising of 28,745 breast cancer cases and 102,997 controls. Subsequently, we developed a breast cancer risk prediction model that integrated breast cancer predisposition genes, a PRS, an epidemiologic risk score (ERS) for non-modifiable risk factors, and another ERS for modifiable risk factors, to examine the joint effects of these factors on breast cancer risk. Our findings demonstrated that an ERS, when combined with the PRS, could contribute to risk stratification for women who have a rare PV in a high-penetrance breast cancer susceptibility gene. The integration of rare high-penetrant and common low-penetrant genetic variation with epidemiologic risk factors into breast cancer risk prediction models has the potential to inform personalized screening and prevention strategies.Next, we assessed the utilization of recommended risk management options following genetic testing for hereditary cancer syndromes based on data from 680 patients who were part of the Cancer Health Assessments Reaching Many (CHARM) study, a multimodal cancer genetics services delivery program. We identified post-testing screening and surgical procedures using electronic health records, and examined utilization in participants who did and did not receive actionable risk management recommendations from study genetic counselors following national guidelines. Our results indicate that implementing CHARM’s risk assessment and testing model increased access to evidence-based genetic services and provided opportunities for patients to engage in recommended preventive care, without encouraging excessive use of risk management strategies.The findings of this dissertation could be useful to identify population subgroups who are especially susceptible to developing breast cancer and help establish personalized risk prevention recommendations for high-risk women.

Age-varying genetic and environmental impact on Breast Cancer

이채은 서울대학교 대학원 2024 국내석사

Breast cancer is a polygenic disease influenced by both genetic and environmental factors. Although numerous studies have examined the genetic information and breast cancer diagnoses across different ancestries and age groups, there is limited research addressing the varying incidence rates based on various factors. A comprehensive understanding of the relationship between breast cancer, age, and various contributing factors is crucial for developing effective strategies for early detection and prevention. Our study statistically evaluates genetic and lifestyle factors influencing breast cancer development across ancestries and age groups of 10-year cohorts, ranging from 20 to 79. We compared European ancestry groups using White British samples from UK Biobank and Asian ancestry groups using Korean samples from Korean Genome and Epidemiology Study (KoGES). The European ancestry group exhibited a significant risk of breast cancer on common genetic variant across the age groups from their 30s to 70s, with the highest risk in the 30s age group (OR=4.19, 95% CI 3.07-5.72, p-value = 2.95E-18) and the lowest risk observed in the 60s age group (OR=2.58, 95% CI 2.30-2.90, p-value = 6.14E-54). For the Asian ancestry group, common variant mutations posed a high risk in individuals in their 40s (OR=1.74, 95% CI 1.16-2.61, p-value = 0.0092). For rare variants, BRCA2 shows the highest risk in 30s age group (OR=12.45, 95% CI 2.49-62.29, p-value = 0.0007), BRCA1 (OR=1.82, 95% CI 1.17-2.81, p-value = 0.01) and CHEK2 (OR=2.31, 95% CI 1.24-4.28, p-value = 0.0115) in 40s age group, indicating early onset on breast cancer. Regarding obesity and lifestyle factors, in the European ancestry group, the 60s age group showed an increased breast cancer risk associated with obesity (OR = 1.29, 95% CI 1.11-1.49, p-value = 0.0008) and alcohol consumption (OR = 1.23, 95% CI 1.04-1.44, p-value = 0.0141). Conversely, physical activity was found to reduce the risk (OR = 0.73, 95% CI 0.59-0.91, p-value = 0.0042) within the same age group. In the Asian group, the highest risk of alcohol consumption was observed in their 40s (OR=2.49, 95% CI 1.28-4.84, p-value = 0.0096), and physical activity was associated with the lowest risk in their 50s (OR=0.58, 95% CI 0.38-0.88, p-value = 0.0101). In this study, the analysis of genetic impact, including both common and rare variants, reveals that being under 50 years of age is usually a significant factor in the onset of breast cancer. Additionally, age is also identified as a critical factor in the onset of breast cancer from an environmental perspective, with varying trends observed across different factors. Future studies are warranted to explore the rare variant data in the Asian ancestry group if such data become available in the future, as this could lead to a better understanding of additional causes of breast cancer. 유방암의 다양한 원인을 이해하는 것은 조기 발견과 적절한 치료를 통해 예방하는 데 중요합니다. 기존 연구들은 다양한 인구와 연령대를 대상으로 유전 정보와 여러 질병의 진단을 조사했지만, 각 연령대의 유방암 위험을 평가하는 연구가 필요해 보였습니다. 본 연구는 유럽계와 아시아계 인구를 대상으로 유전 및 생활 습관이 유방암 발병에 미치는 영향을 연령대별로 분석하고자 하였습니다. 20세에서 79세까지를 10세 단위로 나누어 분석하였으며, 유럽계 인구는 UK Biobank 데이터를, 아시아계 인구는 KoGES 데이터를 사용하였습니다. 유전적 요인 측면에서, 유럽계 인구는 30대부터 70대까지 흔한 변이와 관련된 높은 위험을 보였으며, 30대(OR=4.19, p-value=2.95E-18)에서 가장 높았습니다. 아시아계 인구에서는 40대(OR=1.74, p-value=0.0092)에서 흔한 변이와 관련된 높은 위험이 관찰되었습니다. 희귀 변이에 대한 분석에서, BRCA2는 유럽계 30대에서, BRCA1과 CHEK2는 40대에서 높은 위험을 나타냈습니다. 생활 습관 측면에서는, 유럽계 인구의 경우 60대에서 비만(OR = 1.29, p-value=0.0008)과 음주(OR = 1.23, p-value=0.0141)가 유방암 위험을 증가시키는 요인으로 작용했으며, 신체 활동은 위험 감소(OR = 0.73, p-value=0.0042)와 관련이 있었습니다. 아시아계 인구에서는 40대의 음주(OR=2.49, p-value=0.0096)와 50대의 신체 활동(OR=0.58, p-value=0.0101)에서 각각 높은 위험과 낮은 위험이 관찰되었습니다.

Inflammatory Mediators of Type 2 Diabetes Risk

Downie, Carolina Grace The University of North Carolina at Chapel Hill Pr 2023 해외박사(DDOD)

Inflammation is a pathological feature underlying many chronic diseases, including type 2 diabetes (T2D). Inflammatory mediators include both signaling proteins (e.g., cytokines) and bioactive lipids such as oxylipins, which are derived from oxygenation of fatty acids through cascading pathways mediated by cytochrome P450 (CYP), lipoxygenase (LOX), and cyclooxygenase (COX) enzymes, or through other enzymatic and non-enzymatic reactions. Consequently, there is growing interest in developing oxylipin-targeted therapies for T2D and other cardiometabolic diseases. Enabled by recent advancements in the detection and annotation of oxylipins and related fatty acids, the objective of this study was to characterize the genetic architecture of oxylipins and related fatty acids and evaluate associations between oxylipins and fatty acids and diabetes. We conducted this study in the ancestrally diverse Hispanic Community Health Study/Study of Latinos, a population facing a high burden of diabetes, performing a genome wide association study and observational and innovative causal inference analyses.Our analyses revealed both shared and distinct genetic architecture for n=81 oxylipins and fatty acids of modest to moderate heritability, including genetic loci involved in both oxylipin and fatty acid biosynthesis (e.g., FADS, CYP loci) as well as downstream effects such as small molecule transport (e.g., SLCO1B1, UGT2B7). Moreover, joint two-degree-of-freedom tests extending our models to account for interactions with waist circumference identified an additional locus with a rare top variant that may have anti-inflammatory effects. Consistent with previous research, we identified 46 oxylipins and fatty acids significantly associated with prevalent diabetes, most of which were positively associated with diabetes. Moreover, using our GWAS results as genetic instruments and utilizing a Bayesian model averaging multivariable Mendelian randomization approach accounting for highly correlated exposures, we prioritized heptadecanoic acid (m/z=269.25) as the top causal candidate. Our work demonstrates how metabolomic, genetic, and other population-based phenotypic data can be leveraged to advance understanding of the biological pathways influencing oxylipins and fatty acids, and the role of these bioactive lipids in diabetes. Continued efforts to characterize the genetic factors influencing oxylipins and fatty acids and their role in diabetes risk and progression may help inform the development of preventative or therapeutic drug targets.