Linking allelic variants to variation for complex traits has been a major focus in modern genetics. However, the ability to assess and predict how genetic factors influence complex traits (including human disease) requires an understanding of the specific genes that influence a trait, but also a broader understanding of the genetic architecture of complex traits. Epistatic interactions are crucial when mapping genotypic to phenotypic effects, as mutations in two (or more) genes can produce a phenotype that differs substantially from the expectation of the sum of individual effects. Epistatic interactions are both common and vary considerably in magnitude. Yet little current research focuses on identifying and predicting when mutations will have epistatic interactions and the extent of such effects. I examined the extent of epistatic interactions as a function of individual allelic effects (magnitudes) and wild-type genetic background using the Drosophila melanogaster wing as a model system. The aim of this research is to demonstrate whether individual mutational effects are predictive of the magnitude and direction of epistatic interactions. My results indicate 1) relationships between the average mutant effect and the resulting epistasis and 2) the genetic background can have a strong influence on epistasis. / Thesis / Master of Science (MSc)

Naturally evolving proteins must navigate a vast set of possible sequences to evolve new functions. This process depends on the genotype-phenotype map. Much effort has been directed at measuring protein genotype phenotype maps to uncover evolutionary trajectories that lead to new functions. Often, these maps are too large to comprehensively measure. Sparsely measured maps, however, are prone to missing key evolutionary trajectories. Many groups turn to computational models to infer missing phenotypes. These models treat mutations as independent perturbations to the genotype-phenotype map. A key question is how to handle non-independent effects known as epistasis. In this dissertation, we address two sources of epistasis: 1) global and 2) local epistasis. We find that incorporating global epistasis improves our predictive power, while local epistasis does not. We use our model to infer unknown phenotypes in the Plasmodium falciparum chloroquine transporter (PfCRT) genotype-phenotype map, a protein responsible for conferring drug resistance in malaria. From these predictions, we uncover key evolutionary trajectories that led high resistance. This dissertation includes previously published and unpublished co-authored material. / 2020-01-11

In a previous study, three determinants of fitness were identified as mutant alleles (each designated "e") that arose in yeast populations propagated in divergent environments. In a low-glucose environment, MDS3e and MKT1e interacted positively to confer a fitness advantage. PMA1e from a high-salt environment interacted negatively with MKT1e in low glucose, indicating a mechanism of reproductive isolation. In this thesis, I demonstrated that the negative interaction between PMA1e and MKT1e is mediated by alteration in intracellular pH and likely by a delay of the cell division cycle, while the positive interaction between MDS3e and MKT1e is mediated by changes in gene expression affecting glucose transporter genes. I also confirmed the evolutionary significance of the positive interaction by showing that an MDS3e genetic background is required for the recapitulation of the MKT1e mutation. Collectively, these results illustrate how epistasis can play a central role in both adaptation and speciation.

In a previous study, three determinants of fitness were identified as mutant alleles (each designated "e") that arose in yeast populations propagated in divergent environments. In a low-glucose environment, MDS3e and MKT1e interacted positively to confer a fitness advantage. PMA1e from a high-salt environment interacted negatively with MKT1e in low glucose, indicating a mechanism of reproductive isolation. In this thesis, I demonstrated that the negative interaction between PMA1e and MKT1e is mediated by alteration in intracellular pH and likely by a delay of the cell division cycle, while the positive interaction between MDS3e and MKT1e is mediated by changes in gene expression affecting glucose transporter genes. I also confirmed the evolutionary significance of the positive interaction by showing that an MDS3e genetic background is required for the recapitulation of the MKT1e mutation. Collectively, these results illustrate how epistasis can play a central role in both adaptation and speciation.

In systems biology and genomics, epistasis characters the impact that a substitution at a particular location in a genome can have on a substitution at another location. This phenomenon is often implicated in the evolution of drug resistance or to explain why particular ‘disease-causing’ mutations do not have the same outcome in all in- dividuals. Hence, uncovering these mutations and their locations in a genome is a central question in biology. However, epistasis is notoriously difficult to uncover, es- pecially in fast-evolving organisms. Here, we present a novel statistical approach that takes inspiration from a model developed in ecology and that we adapt to analyze genetic data in a typically fast-evolving system: the influenza A virus. We validate the approach using experimentally-validated data: known interactions are recovered. We further evaluate the ability of our approach to detect epistasis during antigenic shifts or at the emergence of drug resistance. We show that in all cases, epistasis is prevalent in influenza A viruses, involving many pairs of sites linked together in chains, a hallmark of historical contingency. Strikingly, interacting sites are sepa- rated by large physical distances, which entail either long-range structural effects or functional tradeoffs, for which we find support with the emergence of drug resistance. Our work paves a new way for the unbiased detection of epistasis in a wide range of organisms by performing whole-genome scans.

Understanding how genetic modifications, individual or in combination, affect organismal fitness or other phenotypes is a challenge common to several areas of biology, including human health & genetics, metabolic engineering, and evolutionary biology. The importance of a gene can be quantified by measuring the phenotypic impact of its associated genetic perturbations "here and now", e.g. the growth rate of a mutant microbe. However, each gene also maintains a historical record of its cumulative importance maintained throughout millions of years of natural selection in the form of its degree of sequence conservation along phylogenetic branches. This thesis focuses on whether and how the phenotypic and evolutionary importance of genes are related to each other. Towards this goal, I developed a new approach for characterizing the phenotypic consequences of genetic modifications in genome-scale biochemical networks using constraint-based computational models of metabolism. In particular, I investigated the impact of gene loss events on fitness in the model organism Saccharomyces cerevisiae, and found that my new metric for estimating the cost of gene deletion correlates with gene evolutionary rate. I found that previous failures to uncover this correlation using similar techniques may have been the result of an incorrect assumption about how isoenzymes deletions affect the reaction they catalyze. I next hypothesized that the improvement my metric showed in predicting the cost of isoenzyme loss could translate into an improved capacity to predict the impact of pairs of gene deletions involving isoenzymes. Studies of such pair-wise genetic perturbations are important, because the extent to which a genetic perturbation modifies any given phenotype is often dependent on the genetic background upon which it has been performed. This lack of independence within sets of perturbations is termed epistasis. My results showed that, indeed, the new metric displays an increased capacity to predict epistatic interactions between pairs of genes. In addition to shedding light on the relationship between the functional and evolutionary importance of genes, further developments of our approach may lead to better prediction of gene knockout phenotypes, with applications ranging from metabolic engineering to the search for gene targets for therapeutic applications.

No description available.

Alcohol dependence is a complex disease involving polygenes, environment and their interactions. Inadequate consideration of these interactions may have hampered the progress on genome-wide association studies of alcohol dependence. By using the dataset of the Study of Addiction: Genetics and Environment with 3838 subjects, we conducted a genome-wide association studies of alcohol dependence symptom count (ADSC) with a full genetic model considering additive, dominance, epistasis and their interactions with ethnicity, as well as conditions of co-morbid substance dependence. Twenty quantitative trait single nucleotide polymorphisms (QTSs) showed highly significant associations with ADSC, including four previously reported genes (ADH1C, PKNOX2, CPE and KCNB2) and the reported intergenic rs1363605, supporting the overall validity of the analysis. Two QTSs within or near ADH1C showed very strong association in a dominance inheritance mode and increased the phenotype value of ADSC when the effect of co-morbid opiate or marijuana dependence was controlled. Highly significant association was also identified in variants within four novel genes (RGS6, FMN1, NRM and BPTF), two non-coding RNA and two epistasis loci. QTS rs7616413, located near PTPRG encoding a protein tyrosine phosphatase receptor, interacted with rs10090742 within ANGPT1 encoding a protein tyrosine phosphatase in an additive × additive or dominance × additive manner. The detected QTSs contributed to about 20 percent of total heritability, in which dominance and epistasis effects accounted for over 50 percent. These results demonstrated that perturbations arising from gene–gene interaction and conditions of co-morbidity substantially influence the genetic architecture of complex trait.

The genetic requirements for Influenza virus to infect and adapt to new species is largely unknown. To understand the evolutionary steps required by a virus to become virulent, a human virus (A/HK/1/68) (HK), avirulent in mice, was subjected to 20 and 21 serial lung-to-lung passages in mouse. Sequence analysis revealed the emergence of eleven mutations within the NS1 gene of the new virulent strains, many of which occurred in binding sites for transcriptional and translational cellular factors. In the present study we have rescued viruses containing each of the NS1 mouse adapted mutations onto A/PR/8/34 (PR8) backbone. We found 9 of 16 NS1 mutants were adaptive by inducing mortality, body weight loss in BALB/c mice and enhanced virus replication in MDCK cells with properties of host cell interferon transcription inhibition. Sequence comparisons with the highly pathogenic A/Hong Kong/156/1997 (H5N1) and the most severe pandemic A/Brevig Mission/1/1918 (H1N1) NS1 genes showed convergent evolution with some of the mouse adapted viruses for F103L plus M106I and V226I plus R227K mutations respectively. The F103L and M106I mutations in the HK NS1 gene were shown to be adaptive by assessment with respect to replication, early viral protein synthesis, interferon-β antagonism and tropism in the mouse lung. We extended the study and proved increased virulence associated with F103L+M106I mutations in their respective H5N1 NS1 gene on the PR8 and HK backbones, as well as the PR8 NS1 gene and the H9N2 (A/Ck/Bj/1/95) gene in the PR8 and A/WSN/33 backbones respectively. However the V226I and R227K mutations in their respective HK and 1918 NS1 genes slightly enhanced virulence and viral growth at later stages of infection. This study demonstrates that NS1 is a virulence factor; involved in multiple viral processes including interferon antagonism and viral protein synthesis. Furthermore, NS1 mutations acquired during mouse adaptation are proven to be adaptive in human, mouse and avian NS1 genes.

Psychotic disorders such as schizophrenia and bipolar disorder have a strong genetic component. The aetiology of psychoses is known to be complex, including additive effects from multiple susceptibility genes, interactions between genes, environmental risk factors, and gene by environment interactions. With the development of new technologies such as genome-wide association studies and imputation of ungenotyped variants, the amount of genomic data has increased dramatically leading to the necessary use of Machine Learning techniques. Random Forest has been widely used to study the underlying genetic factors of psychiatric disorders such as epistasis and gene-gene interactions. Several authors have investigated the ability of this algorithm in finding single and interaction effects, but have reported contradictory results. Therefore, in order to examine Random Forest ability of detecting single and interaction effects based on different variable importance measures, I conducted a simulation study assessing whether the algorithm was able to detect single and interaction models under different correlation conditions. The results suggest that the optimal Variable Importance Measures to use in real situations under correlation is the unconditional unscaled permutation variable importance measure. Several studies have shown bias in one of the most popular variable importance measures, the Gini importance. Hence, in a second simulation study I study whether the Gini variable importance is influenced by the variability of predictors, the precision of measuring them, and the variability of the error. Evidence of other biases in this variable importance was found. The results from the first simulation study were used to study whether genes related to 29 molecular biomarkers, which have been associated with schizophrenia, influence risk for schizophrenia in a case-control study of 26476 cases and 31804 controls from 39 different European ancestry cohorts. Single effects from ACAT2 and TNC genes were detected to contribute risk for schizophrenia. ACAT2 is a gene in the chromosome 6 which is related to energy metabolism. Transcriptional differences have been shown in schizophrenia brain tissue studies. TNC is expressed in the brain where is involved in the migration of the neurons and axons. In addition, we also used the simulation results to examine whether interactions between genes associated with abnormal emotion/affect behaviour influence risk for psychosis and cognition in humans, in a case-control study of 2049 cases and 1794 controls. Before correcting for multiple testing, significant interactions between CRHR1 and ESR1, and between MAPT and ESR1, and among CRHR1, ESR1 and TOM1L2, and among MAPT, ESR1 and TOM1L2 were observed in abnormal fear/anxiety-related behaviour pathway. There was no evidence for epistasis after Bonferroni correction.