Greg Gibson

Emory University, Atlanta, Georgia, United States

Are you Greg Gibson?

Claim your profile

Publications (134)1201.44 Total impact

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Continued exposure to malaria-causing parasites in endemic regions of malaria induces significant levels of acquired immunity in adult individuals. A better understanding of the transcriptional basis for this acquired immunological response may provide insight into how the immune system can be boosted during vaccination, and into why infected individuals differ in symptomology. Peripheral blood gene expression profiles of 9 semi-immune volunteers from a Plasmodium vivax malaria prevalent region (Buenaventura, Colombia) were compared to those of 7 naïve individuals from a region with no reported transmission of malaria (Cali, Colombia) after a controlled infection mosquito bite challenge with P. vivax. A Fluidigm nanoscale quantitative RT-PCR array was used to survey altered expression of 96 blood informative transcripts at 7 timepoints after controlled infection, and RNASeq was used to contrast pre-infection and early parasitemia timepoints. There was no evidence for transcriptional changes prior to the appearance of blood stage parasites at day 12 or 13, at which time there was a strong interferon response and, unexpectedly, down-regulation of transcripts related to inflammation and innate immunity. This differential expression was confirmed with RNASeq, which also suggested perturbations of aspects of T cell function and erythropoiesis. Despite differences in clinical symptoms between the semi-immune and malaria naïve individuals, only subtle differences in their transcriptomes were observed, although 175 genes showed significantly greater induction or repression in the naïve volunteers from Cali. Gene expression profiling of whole blood reveals the type and duration of the immune response to P. vivax infection, and highlights a subset of genes that may mediate adaptive immunity.
    PLoS Neglected Tropical Diseases 08/2015; 9(8):e0003978. DOI:10.1371/journal.pntd.0003978 · 4.49 Impact Factor
  • Source
    Greg Gibson · Joseph E Powell · Urko M Marigorta
    [Show abstract] [Hide abstract]
    ABSTRACT: Expression quantitative trait locus analysis has emerged as an important component of efforts to understand how genetic polymorphisms influence disease risk and is poised to make contributions to translational medicine. Here we review how expression quantitative trait locus analysis is aiding the identification of which gene(s) within regions of association are causal for a disease or phenotypic trait; the narrowing down of the cell types or regulators involved in the etiology of disease; the characterization of drivers and modifiers of cancer; and our understanding of how different environments and cellular contexts can modify gene expression. We also introduce the concept of transcriptional risk scores as a means of refining estimates of individual liability to disease based on targeted profiling of the transcripts that are regulated by polymorphisms jointly associated with disease and gene expression.
    Genome Medicine 06/2015; 7(1):60. DOI:10.1186/s13073-015-0186-7 · 4.94 Impact Factor
  • Cytotherapy 06/2015; 17(6):S43. DOI:10.1016/j.jcyt.2015.03.457 · 3.10 Impact Factor
  • Greg Gibson
    Science 05/2015; 348(6235):640-1. DOI:10.1126/science.aab3002 · 31.48 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Histiocytoid cardiomyopathy (Histiocytoid CM) is a rare form of cardiomyopathy observed predominantly in newborn females that is fatal unless treated early in life. We have performed whole exome sequencing on five parent-proband trios and identified nuclear-encoded mitochondrial protein mutations in three cases. The molecular genetic basis of Histiocytoid CM remains unknown despite several hypotheses in medical literature. The findings presented in this manuscript may represent components of genetic etiologies for this heterogeneous disease. Two probands had de novo non-sense mutations in the second exon of the X-linked nuclear gene NDUFB11. A third proband was doubly heterozygous for inherited rare variants in additional components of complex I, NDUFAF2 and NDUFB9, confirming that Histiocytoid CM is genetically heterogeneous. In a fourth case, the proband with Histiocytoid CM inherited a mitochondrial mutation from her heteroplasmic mother, as did her brother who presented with cardiac arrhythmia. Strong candidate recessive or compound heterozygous variants were not found for this individual or for the fifth case. Although NDUFB11 has not been implicated before in cardiac pathology, morpholino-mediated knockdown of ndufb11 in zebrafish embryos generated defective cardiac tissue with cardiomegaly, looping defects, and arrhythmia which suggests the role of NDUFB11 in the pathogenesis of this abnormal cardiac pathology. Taken together, the unbiased whole exome sequencing approach confirms the suspected genetic heterogeneity of Histiocytoid CM. Therefore, the novel NDUFB11 mutation may cause a complex 1 deficiency in synergy with additional unknown mtDNA variants. © 2015 Wiley Periodicals, Inc. © 2015 Wiley Periodicals, Inc.
    American Journal of Medical Genetics Part A 04/2015; DOI:10.1002/ajmg.a.37138 · 2.05 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Bacteria colonize cystic fibrosis (CF) airways, and although T cells with appropriate Ag specificity are present in draining lymph nodes, they are conspicuously absent from the lumen. To account for this absence, we hypothesized that polymorphonuclear neutrophils (PMNs), recruited massively into the CF airway lumen and actively exocytosing primary granules, also suppress T cell function therein. Programmed death-ligand 1 (PD-L1), which exerts T cell suppression at a late step, was expressed bimodally on CF airway PMNs, delineating PD-L1(hi) and PD-L1(lo) subsets, whereas healthy control (HC) airway PMNs were uniformly PD-L1(hi). Blood PMNs incubated in CF airway fluid lost PD-L1 over time; in coculture, Ab blockade of PD-L1 failed to inhibit the suppression of T cell proliferation by CF airway PMNs. In contrast with PD-L1, arginase 1 (Arg1), which exerts T cell suppression at an early step, was uniformly high on CF and HC airway PMNs. However, arginase activity was high in CF airway fluid and minimal in HC airway fluid, consistent with the fact that Arg1 activation requires primary granule exocytosis, which occurs in CF, but not HC, airway PMNs. In addition, Arg1 expression on CF airway PMNs correlated negatively with lung function and positively with arginase activity in CF airway fluid. Finally, combined treatment with arginase inhibitor and arginine rescued the suppression of T cell proliferation by CF airway fluid. Thus, Arg1 and PD-L1 are dynamically modulated upon PMN migration into human airways, and, Arg1, but not PD-L1, contributes to early PMN-driven T cell suppression in CF, likely hampering resolution of infection and inflammation. Copyright © 2015 by The American Association of Immunologists, Inc.
    The Journal of Immunology 04/2015; 194:0-0. DOI:10.4049/jimmunol.1500312 · 5.36 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Autologous bone marrow derived mesenchymal stromal cells (MSCs) for adoptive cell therapy of luminal Crohn's Disease (CD) are being tested in clinical trials. However, CD is associated with dysregulation of autophagy and its effect on MSC's immunobiology is unknown. Here we demonstrate no quantitative difference in phenotype, in vitro growth kinetics and molecular signatures to IFNγ between MSCs derived from CD and healthy individuals. CD MSCs were indistinguishable from those derived from healthy controls at inhibiting T cell proliferation through an indoleamine 2,3-dioxygenase (IDO) dependent mechanism. Upon IFNγ pre-licensing, both MSC populations inhibit T cell effector functions. Neither a single nucleotide polymorphism (SNP) rs7820268 in the IDO gene, nor a widely reported CD predisposing SNP ATG16L1rs2241880 modulated the suppressive function of MSCs carrying these haplotypes. IFNγ stimulation or co-culture with activated T cells upregulated the expression of autophagy genes and/or vacuoles on MSCs. Pharmacological blockade of autophagy pathway did not reverse the immunosuppressive properties and IFNγ responsiveness of MSCs confirming the absence of a functional link between these two cell biochemical properties. We conclude that autophagy, but not IDO and IFNγ responsiveness, is dispensable for MSC's immunosuppressive properties. MSCs from CD subjects are functionally analogous to those of healthy individuals.Molecular Therapy (2015); doi:10.1038/mt.2015.67.
    Molecular Therapy 04/2015; DOI:10.1038/mt.2015.67 · 6.43 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The Center for Health Discovery and Wellbeing (CHDWB) is an academic program designed to evaluate the efficacy of clinical self-knowledge and health partner counseling for development and maintenance of healthy behaviors. This paper reports on the change in health profiles for over 90 traits, measured in 382 participants over three visits in the 12 months following enrolment. Significant changes in the desired direction of improved health are observed for many traits related to cardiovascular health, including BMI, blood pressure, cholesterol, and arterial stiffness, as well as for summary measures of physical and mental health. The changes are most notable for individuals in the upper quartile of baseline risk, many of whom showed a positive correlated response across clinical categories. By contrast, individuals who start with more healthy profiles do not generally show significant improvements and only a modest impact of targeting specific health attributes was observed. Overall, the CHDWB model shows promise as an effective intervention particularly for individuals at high risk for cardiovascular disease.
    12/2014; 4(4):489-507. DOI:10.3390/jpm4040489
  • Thanawadee Preeprem · Greg Gibson
    [Show abstract] [Hide abstract]
    ABSTRACT: The identification of genetic factors that influence drug responses often focuses on genes whose variants are predicted to alter pharmacokinetic and/or pharmacodynamic parameters—leading to an increased risk of drug toxicity or therapeutic failure. In this study, we selected 48 genes identified as “Very Important Pharmacogenes (VIPs)” by the PharmGKB database, and developed a fivefeature Structural Disturbance Score (SDS) for their amino acid variants. “SDS Pharmacogenes” is a score that categorizes distinguishable characteristic profiles that annotate VIP variants as functional rather than neutral mutations. Unlike most existing conservation-based measures, SDS Pharmacogenes can be used to evaluate unknown variants of currently 45/48 VIPs and predict the degree to which each one will have strong impacts towards pharmacogenomics, potentially aiding optimization of drug therapy. SDS Pharmacogenes is built upon a systematic screening for structural disturbance of amino acid mutations within the 45/48 VIPs in the context of their 3-dimensional (3D) protein structures. Our variant evaluation pipeline focuses on the changes in inter-residue bonding, protein stability, protein flexibility, drug binding capability, protein-protein interactions, and amino acid dissimilarity, in addition to the localization of the variants and the amino acid secondary structure preference. While expertise in 3D-protein analysis is beneficial, our implementation does not require that an individual with experience in protein structures be engaged in the personalized genome evaluation, nor expect the users must have bioinformatic backgrounds. In addition, the analysis pipeline is systematic and scalable, thus expected to keep pace with the rapid accumulation of pharmacogenomic data. The SDS Pharmacogenes web application is publicly available at http://sdsp. cig.biology.gatech.edu.
    Current Pharmacogenomics and Personalized Medicine (Formerly Current Pharmacogenomics) 12/2014; 12(4). DOI:10.2174/1875692113666150408222109
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We describe a multi-omic approach to understanding the effects that the anti-malarial drug pyrimethamine has on immune physiology in rhesus macaques (Macaca mulatta). Whole blood and bone marrow (BM) RNA-Seq and plasma metabolome profiles (each with over 15,000 features) have been generated for five naïve individuals at up to seven timepoints before, during and after three rounds of drug administration. Linear modeling and Bayesian network analyses are both considered, alongside investigations of the impact of statistical modeling strategies on biological inference. Individual macaques were found to be a major source of variance for both omic data types, and factoring individuals into subsequent modeling increases power to detect temporal effects. A major component of the whole blood transcriptome follows the BM with a time-delay, while other components of variation are unique to each compartment. We demonstrate that pyrimethamine administration does impact both compartments throughout the experiment, but very limited perturbation of transcript or metabolite abundance was observed following each round of drug exposure. New insights into the mode of action of the drug are presented in the context of pyrimethamine's predicted effect on suppression of cell division and metabolism in the immune system.
    Frontiers in Cell and Developmental Biology 10/2014; 2:54. DOI:10.3389/fcell.2014.00054
  • Source
    Aliza P. Wingo · Greg Gibson
    [Show abstract] [Hide abstract]
    ABSTRACT: Prospective epidemiological studies found that generalized anxiety disorder (GAD) can impair immune function and increase risk for cardiovascular disease or events. Mechanisms underlying the physiological reverberations of anxiety, however, are still elusive. Hence, we aimed to investigate molecular processes mediating effects of anxiety on physical health using blood gene expression profiles of 336 community participants (157 anxious and 179 control). We examined genome-wide differential gene expression in anxiety, as well as associations between nine major modules of co-regulated transcripts in blood gene expression and anxiety. No significant differential expression was observed in women, but 631 genes were differentially expressed between anxious and control men at the false discovery rate of 0.1 after controlling for age, body mass index, race, and batch effect. Gene set enrichment analysis (GSEA) revealed that genes with altered expression levels in anxious men were involved in response of various immune cells to vaccination and to acute viral and bacterial infection, and in a metabolic network affecting traits of metabolic syndrome. Further, we found one set of 260 co-regulated genes to be significantly associated with anxiety in men after controlling for the relevant covariates, and demonstrate its equivalence to a component of the stress-related conserved transcriptional response to adversity profile. Taken together, our results suggest potential molecular pathways that can explain negative effects of GAD observed in epidemiological studies. Remarkably, even mild anxiety, which most of our participants had, was associated with observable changes in immune-related gene expression levels. Our findings generate hypotheses and provide incremental insights into molecular mechanisms mediating negative physiological effects of GAD.
    Brain Behavior and Immunity 10/2014; 43. DOI:10.1016/j.bbi.2014.09.016 · 6.13 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Replying to A. R. Wood et al. 514, http://dx.doi.org/10.1038/nature13691 (2014).We thank Wood et al. for their interesting observations and although their proposed mechanism does not explain all our reported results, we acknowledge that alternative mechanisms could be behind the observation of epistatic signals. Although we replicate our results in large, independent samples, 19/30 of our reported interactions (Table 1 in ref. 2), Wood et al. do not replicate in the InCHIANTI data set (n = 450) at a type-I error rate of 0.05/30 = 0.002, including none of our reported cis-trans interactions. Having insufficient data to replicate the discovery interactions makes it problematic to draw firm conclusions on the reported cis-trans effects.
    Nature 10/2014; 514(7520):E5-6. DOI:10.1038/nature13692 · 42.35 Impact Factor
  • Greg Gibson
    Nature 07/2014; 512(7512). DOI:10.1038/nature13649 · 42.35 Impact Factor
  • Source
    Urko M Marigorta · Greg Gibson
    [Show abstract] [Hide abstract]
    ABSTRACT: The switch to a modern lifestyle in recent decades has coincided with a rapid increase in prevalence of obesity and other diseases. These shifts in prevalence could be explained by the release of genetic susceptibility for disease in the form of gene-by-environment (GxE) interactions. Yet, the detection of interaction effects requires large sample sizes, little replication has been reported, and a few studies have demonstrated environmental effects only after summing the risk of GWAS alleles into genetic risk scores (GRSxE). We performed extensive simulations of a quantitative trait controlled by 2500 causal variants to inspect the feasibility to detect gene-by-environment interactions in the context of GWAS. The simulated individuals were assigned either to an ancestral or a modern setting that alters the phenotype by increasing the effect size by 1.05-2-fold at a varying fraction of perturbed SNPs (from 1 to 20%). We report two main results. First, for a wide range of realistic scenarios, highly significant GRSxE is detected despite the absence of individual genotype GxE evidence at the contributing loci. Second, an increase in phenotypic variance after environmental perturbation reduces the power to discover susceptibility variants by GWAS in mixed cohorts with individuals from both ancestral and modern environments. We conclude that a pervasive presence of gene-by-environment effects can remain hidden even though it contributes to the genetic architecture of complex traits.
    Frontiers in Genetics 07/2014; 5:225. DOI:10.3389/fgene.2014.00225
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Single-cell analysis has the potential to provide us with a host of new knowledge about biological systems, but it comes with the challenge of correctly interpreting the biological information. While emerging techniques have made it possible to measure inter-cellular variability at the transcriptome level, no consensus yet exists on the most appropriate method of data analysis of such single cell data. Methods for analysis of transcriptional data at the population level are well established but are not well suited to single cell analysis due to their dependence on population averages. In order to address this question, we have systematically tested combinations of methods for primary data analysis on single cell transcription data generated from two types of primary immune cells, neutrophils and T lymphocytes. Cells were obtained from healthy individuals, and single cell transcript expression data was obtained by a combination of single cell sorting and nanoscale quantitative real time PCR (qRT-PCR) for markers of cell type, intracellular signaling, and immune functionality. Gene expression analysis was focused on hierarchical clustering to determine the existence of cellular subgroups within the populations. Nine combinations of criteria for data exclusion and normalization were tested and evaluated. Bimodality in gene expression indicated the presence of cellular subgroups which were also revealed by data clustering. We observed evidence for two clearly defined cellular subtypes in the neutrophil populations and at least two in the T lymphocyte populations. When normalizing the data by different methods, we observed varying outcomes with corresponding interpretations of the biological characteristics of the cell populations. Normalization of the data by linear standardization taking into account technical effects such as plate effects, resulted in interpretations that most closely matched biological expectations. Single cell transcription profiling provides evidence of cellular subclasses in neutrophils and leukocytes that may be independent of traditional classifications based on cell surface markers. The choice of primary data analysis method had a substantial effect on the interpretation of the data. Adjustment for technical effects is critical to prevent misinterpretation of single cell transcript data.
    PeerJ 06/2014; 2:e452. DOI:10.7717/peerj.452 · 2.10 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Craniosynostosis, the premature fusion of one or more skull sutures, occurs in approximately 1 in 2500 infants, with the majority of cases non-syndromic and of unknown etiology. Two common reasons proposed for premature suture fusion are abnormal compression forces on the skull and rare genetic abnormalities. Our goal was to evaluate whether different sub-classes of disease can be identified based on total gene expression profiles. RNA-Seq data were obtained from 31 human osteoblast cultures derived from bone biopsy samples collected between 2009 and 2011, representing 23 craniosynostosis fusions and 8 normal cranial bones or long bones. No differentiation between regions of the skull was detected, but variance component analysis of gene expression patterns nevertheless supports transcriptome-based classification of craniosynostosis. Cluster analysis showed 4 distinct groups of samples; 1 predominantly normal and 3 craniosynostosis subtypes. Similar constellations of sub-types were also observed upon re-analysis of a similar dataset of 199 calvarial osteoblast cultures. Annotation of gene function of differentially expressed transcripts strongly implicates physiological differences with respect to cell cycle and cell death, stromal cell differentiation, extracellular matrix (ECM) components, and ribosomal activity. Based on these results, we propose non-syndromic craniosynostosis cases can be classified by differences in their gene expression patterns and that these may provide targets for future clinical intervention.
    06/2014; 2:121-130. DOI:10.7150/jgen.8833
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Background Genetic risk scores have been developed for coronary artery disease and atherosclerosis, but are not predictive of adverse cardiovascular events. We asked whether peripheral blood expression profiles may be predictive of acute myocardial infarction (AMI) and/or cardiovascular death. Methods Peripheral blood samples from 338 subjects aged 62 ± 11 years with coronary artery disease (CAD) were analyzed in two phases (discovery N = 175, and replication N = 163), and followed for a mean 2.4 years for cardiovascular death. Gene expression was measured on Illumina HT-12 microarrays with two different normalization procedures to control technical and biological covariates. Whole genome genotyping was used to support comparative genome-wide association studies of gene expression. Analysis of variance was combined with receiver operating curve and survival analysis to define a transcriptional signature of cardiovascular death. Results In both phases, there was significant differential expression between healthy and AMI groups with overall down-regulation of genes involved in T-lymphocyte signaling and up-regulation of inflammatory genes. Expression quantitative trait loci analysis provided evidence for altered local genetic regulation of transcript abundance in AMI samples. On follow-up there were 31 cardiovascular deaths. A principal component (PC1) score capturing covariance of 238 genes that were differentially expressed between deceased and survivors in the discovery phase significantly predicted risk of cardiovascular death in the replication and combined samples (hazard ratio = 8.5, P < 0.0001) and improved the C-statistic (area under the curve 0.82 to 0.91, P = 0.03) after adjustment for traditional covariates. Conclusions A specific blood gene expression profile is associated with a significant risk of death in Caucasian subjects with CAD. This comprises a subset of transcripts that are also altered in expression during acute myocardial infarction.
    Genome Medicine 05/2014; 6(5):40. DOI:10.1186/gm560 · 4.94 Impact Factor
  • Source
    Thanawadee Preeprem · Greg Gibson
    [Show abstract] [Hide abstract]
    ABSTRACT: We have developed a novel structure-based evaluation for missense variants that explicitly models protein structure and amino acid properties to predict the likelihood that a variant disrupts protein function. A structural disruption score (SDS) is introduced as a measure to depict the likelihood that a case variant is functional. The score is constructed using characteristics that distinguish between causal and neutral variants within a group of proteins. The SDS score is correlated with standard sequence-based deleteriousness, but shows promise for improving discrimination between neutral and causal variants at less conserved sites. The prediction was performed on 3-dimentional structures of 57 gene products whose homozygous SNPs were identified as case-exclusive variants in an exome sequencing study of epilepsy disorders. We contrasted the candidate epilepsy variants with scores for likely benign variants found in the EVS database, and for positive control variants in the same genes that are suspected to promote a range of diseases. To derive a characteristic profile of damaging SNPs, we transformed continuous scores into categorical variables based on the score distribution of each measurement, collected from all possible SNPs in this protein set, where extreme measures were assumed to be deleterious. A second epilepsy dataset was used to replicate the findings. Causal variants tend to receive higher sequence-based deleterious scores, induce larger physico-chemical changes between amino acid pairs, locate in protein domains, buried sites or on conserved protein surface clusters, and cause protein destabilization, relative to negative controls. These measures were agglomerated for each variant. A list of nine high-priority putative functional variants for epilepsy was generated. Our newly developed SDS protocol facilitates SNP prioritization for experimental validation.
    Frontiers in Genetics 04/2014; 5:82. DOI:10.3389/fgene.2014.00082
  • Source
    Journal of the American College of Cardiology 04/2014; 63(12):A14. DOI:10.1016/S0735-1097(14)60014-0 · 15.34 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Systems biology is an approach to dissection of complex traits that explicitly recognizes the impact of genetic, physiological and environmental interactions in the generation of phenotypic variation. We describe comprehensive transcriptional and metabolic profiling in Drosophila melanogaster across four diets, finding little overlap in modular architecture. Genotype and genotype-by-diet interactions are a major component of transcriptional variation (24% and 5.3% of the total variation respectively) while there were no main effects of diet (<1%). Genotype was also a major contributor to metabolomic variation (16%), but in contrast to the transcriptome, diet had a large effect (9%) and the interaction effect was minor (2%) for the metabolome. Yet specific principal components of these molecular phenotypes measured in larvae are strongly correlated with particular Metabolic Syndrome-like phenotypes such as pupal weight, larval sugar content and triglyceride content, development time, and cardiac arrhythmia in adults. The second principal component of the metabolomic profile is especially informative across these traits with glycine identified as a key loading variable. To further relate this physiological variability to genotypic polymorphism, we performed evolve-and-resequence experiments, finding rapid and replicated changes in gene frequency across hundreds of loci that are specific to each diet. Adaptation to diet is thus highly polygenic. However, loci differentially transcribed across diet or previously identified by RNAi knockdown or expression QTL analysis where not the loci responding to dietary selection. Therefore, loci that respond to the selective pressures of diet cannot be readily predicted a priori from functional analyses.
    Genetics 03/2014; 197(2). DOI:10.1534/genetics.114.163857 · 4.87 Impact Factor

Publication Stats

9k Citations
1,201.44 Total Impact Points

Institutions

  • 2015
    • Emory University
      • School of Medicine
      Atlanta, Georgia, United States
  • 2010–2015
    • Georgia Institute of Technology
      • School of Biology
      Atlanta, Georgia, United States
  • 1999–2011
    • North Carolina State University
      • • Department of Genetics
      • • Bioinformatics Research Center
      Raleigh, NC, United States
  • 2008–2009
    • University of Queensland
      • Queensland Brain Institute
      Brisbane, Queensland, Australia
  • 2007
    • SAS Institute
      North Carolina, United States
  • 2005
    • Air University
      Maxwell, California, United States
  • 2004
    • University of California, Davis
      • Department of Evolution & Ecology
      Davis, California, United States
    • Cornell University
      • Department of Molecular Biology and Genetics
      Итак, New York, United States
  • 1996–2000
    • University of Michigan
      • Department of Biology
      Ann Arbor, Michigan, United States