This dissertation tries to answer some important questions in labor and transportation economics in an attempt to broaden understanding of different topics.In the first chapter “Correlated Labor Market Risk and Housing Investment” I show that households have lower levels of housing investment when they live in areas with labor markets that are more correlated with their industry of employment. I find that if a household lives in an area where many other households work in the same or similar industries, then housing may be a riskier asset as it is more correlated with labor market income. In response household decrease their investment in housing, and this This decline is driven by concentrations and riskiness of other correlated industries, suggesting agglomeration in one industry can have negative spillovers to workers of other related industries. The second chapter “Loyalty rewards and redemption behavior: Stylized facts for the U.S. airline industry” joint with Alexander Luttmann, provides a novel identification method for frequent flyer tickets in a comment industry database, DB1B. Using the FFAs we identify, we show how the characteristics of award tickets differ from paid tickets and how these characteristics have changed over time. We then demonstrate how various market and product quality characteristics influence the share of passengers traveling on FFAs. Finally, we find that price dispersion increases on routes with larger shares of frequent flyer passengers, implying that airline loyalty programs enhance market power. In the third chapter, “Help Really Wanted? The Impact of Age Stereotypes in Job Ads on Applications from Older Workers” joint with Ian Burn, Daniel Firoozi and David Neumark, We construct job ads for administrative assistant, retail, and security guard jobs, using language from real job ads collected in a prior large-scale correspondence study. We modify the job-ad language to randomly vary whether the job ad includes ageist language regarding age-related stereotypes. In contrast to a correspondence study in which job searchers are artificial and researchers study the responses of real employers, in our research the job ads are artificial and we study the responses of real job searchers. We find that job-ad language related to ageist stereotypes, even when the language is not blatantly or specifically age-related, deters older workers from applying for jobs. The change in the age distribution of applicants is large, with significant declines in the average and median age, the 75th percentile of the age distribution, and the share of applicants over 40. Based on these estimates and those from the correspondence study, and the fact that we use real-world ageist job-ad language, we conclude that job-ad language that deters older workers from applying for jobs can have roughly as large an impact on hiring of older workers as direct age discrimination in hiring.

Coral reefs are the ecological, economic, and social backbone of tropical coastal communities. Yet, more than half of the world’s tropical reefs are gone. With further declines predicted, there is an urgent need to develop effective solutions to resurrect degraded reefs. Part of the challenge of reef restoration is that coral reefs are extremely complex ecosystems. As such, they are often studied via a reductionist approach, whereas rebuilding an ecosystem requires a comprehensive plan to reconstruct the ecological processes necessary for a functioning reef ecosystem. Unfortunately, while we have a mechanistic understanding of many factors driving coral reef decline, there is a dearth of information available to guide us on how to restore degraded coral reefs and recover their ecosystem functions.

To this end, my dissertation examines how harnessing key ecological processes on coral reefs can facilitate the pace and success of coral reef restoration. In Chapter 1, I investigated the role of restored coral density on habitat production, and explored mechanisms contributing to density dependence in coral restoration. I found evidence of a unimodal relationship among restored Acropora cervicornis colonies suggesting positive density dependence at intermediate densities. Importantly, these findings highlight the fundamental role that basic restoration design elements, like outplant density, play in the success or failure of coral restoration. In Chapter 2, I assessed the importance of genotypic identity and diversity in restoration outcomes. Using a field experiment, I identified a tradeoff between thermal-tolerance and growth rates among A. cervicornis genotypes, suggesting genotypic identity is a critical factor to incorporate into restoration planning.

While restoring individuals to rebuild coral populations is an important first step in coral restoration, outplanted corals do not exist in isolation when transplanted to a degraded reef. Non-scleractinian invertebrates like sponges, gorgonians, and zoanthids are increasing on reefs, yet there is a paucity of data on interactions between these increasingly common organisms and corals. In Chapter 3, using observational surveys I found that competitive interactions were pervasive on Florida reefs, with 60% of sessile benthic invertebrates interacting with at least one other invertebrate. Further, results from a common garden competition experiment demonstrated that non-scleractinians like sponges and zoanthids consistently outcompeted the common species Porites porites and Siderastrea siderea, suggesting competition may limit the success of these coral species and is likely to remain an important process structuring contemporary coral reef communities.

Chapter 4 addresses our knowledge gap on the effects of coral restoration on reef communities and important ecosystem functions. To do so, I conducted surveys of sites in the Florida Keys that had undergone coral restoration paired with unmanipulated control sites. I found that coral restoration enhanced coral populations, increasing coral cover 4-fold, but manifested in limited differences in coral and fish communities. Interestingly, damselfishes, whose territorial behavior may deter important processes like herbivory, were the only group of fishes that positively responded to coral restoration. These findings suggest that additional considerations beyond outplanting corals will likely be necessary to effectively restore coral reefs in a time of increasingly frequent and intense disturbances.

Lastly, in Chapter 5 I synthesized literature on coral restoration and reef ecology to identify key drivers of recovery and propose a path forward to improve coral restoration. Specifically, restoration practitioners can manipulate factors such as the density, diversity and identity of transplanted corals and leverage existing ecological processes on coral reefs to restore positive feedback processes, or disrupt negative feedback processes, and facilitate restoration success. Importantly, the results of this dissertation can be directly applied to inform how coral reef restoration is conducted and improve our ability to effectively restore degraded coral reefs.

This dissertation describes a result that compares two level subgroups on different inner forms of GSp(4), and then uses this result and a conjecture of Ibukiyama’s to compute paramodular forms for all prime levels below 400. In the process 78 generic forms were computed, of which 47 had not been previously computed.

Progress in metabolomics has brought the field from investigations of pre-selectedcompound lists and limited sample size toward comprehensive compound exploration of large sample size. This shift in focus demanded corresponding advances in informatics areas that we explore in this dissertation, such as in-silico compound identification tools, metabolomics meta- analysis, and metabolomics repository design. In Chapter 1, we focus on compound identification. Compound identification is traditionally treated as an information retrieval problem, where unknown compounds are identified by comparing their observed signals to the signals of chemical standards. Unfortunately, the metabolome contains significantly more compounds than standards, so there is a desire to computationally expand the space of indexed signals. Here, we benchmark a tool, CFM-ID, that predicts the signal of a compound based on its structure. We show that there is much progress needed in this area by determining that CFM-ID’s predictions could be readily replicated via heuristic rules that focus on structure. Extrapolating these ideas emphasizes the need for increased machine learning model training set sizes and standardization due to the complexity of the physics and statistical mechanics that mass spectrometry signals reflect. In Chapter 2, we focus on meta-analysis of metabolomics studies. We believe that the synchronization of many independent datasets will allow for biological insights of high confidence and/or high generality. To this end, we developed a tool named BinDiscover, which allows for rapid hypothesis generation by enabling user-directed exploration of over 150,000 samples processed at the West Coast Metabolomics Center. We believe that this tool improves existing repository meta-analysis for several reasons. First, it is programmatic in nature, which allows for meta-analysis on a timescale of minutes rather than months. Second, the meta-analysis that it iv enables is focused on sample metadata rather than study hypotheses, which dramatically expands the number of investigations that can be conducted. Third, it is dramatically easier to use than existing options. Finally, it showcases our novel procedure, ontologically-grouped-differential analysis, which allows for the convenient comparison of categories of samples (e.g., mammals digestive system organs vs. bacterial cells) in order to produce tractable amounts high-confidence results. In Chapter 3, we focus on repository design. We strongly believe that enabling the programmatic meta-analysis developed on in Chapter 2 onto a larger-scale, community- contributed repositories of metabolomics data will enable massive clinical progress. To this end, we developed a tool that standardizes sample metadata. At current, user-submitted sample metadata matrices preclude programmatic meta-analysis because they suffer from the looseness and complexity of natural language. Our multistep standardization tool employs machine learning models embedded into an intuitive frontend to ensure that only high-quality sample descriptions are lodged into repositories. Finally, in the appendix, we share several projects spanning the topics of the main chapters. In the first part, we share ClusterBase, which is a computational platform that uses network analysis to organize and annotate spectral data from metabolomics studies. In the second part, we share an automatic compound-ID workflow that harnessed the online CFM-ID tool. Finally, in the third part, we describe a machine learning approach to predicting spectral intensities that can augment quantum mechanically predicted spectra.

RNA interference (RNAi) pathways are critical eukaryotic post-transcriptional gene regulatory systems that control the expression of at least one third of all human coding genes. In each of these pathways, short single-stranded RNAs (ssRNAs) bind to cognate messenger RNAs (mRNAs) and direct their endonucleolytic cleavage or translational repression by RNA-Induced Silencing Complexes (RISCs). In the cytoplasmic portion of these pathways, silencing is triggered by long double-stranded RNA (dsRNA) precursors (pre-siRNAs) or hairpin RNAs (pre-miRNAs), which are processed by the RNase III enzyme Dicer to yield short 21-23 nucleotide (nt) duplex RNAs termed short-interfering RNAs (siRNAs) and microRNAs (miRNAs), respectively. These short duplex RNAs are then loaded onto Argonaute2 (Ago2), the catalytic component of RISC. Ago2 retains one strand of the duplex (the "guide" strand) for subsequent gene targeting and discards the other. This process of asymmetrically loading siRNA or miRNA duplexes onto RISCs such that one strand is preferentially retained is fundamental to the target specificity of RNAi pathways.

Structural studies of the individual components of the RNAi machinery have proven immensely informative to our mechanistic understanding of the individual steps of RNAi pathways such as substrate cleavage by Dicer and subsequent mRNA targeting by Ago2. What has been missing is a clear picture of how different RNAi components functionally interact to transfer small RNAs from one protein to another in a way that ultimately allows for the selection of the proper guide strand by Ago2. In Drosophila and humans, protein complexes called RISC Loading Complexes (RLCs) have been implicated in facilitating this process. At their core, these complexes are composed of Ago2, Dicer, and a dsRNA binding protein (dsRBP). Two such complexes exist in humans and are defined by their associated dsRBP. One complex contains HIV-1 Trans-activation Response (TAR) RNA Binding Protein (TRBP) and the other contains Protein Activator of PKR (PACT). Here we have used electron microscopy (EM) and single particle reconstruction methods to show that human Dicer has an L-shaped architecture consisting of a long catalytic branch and a shorter DExH/D helicase-containing base branch. Further, we present a three-dimensional (3D) reconstruction of a highly dynamic human RLC containing Ago2, Dicer, and TRBP. TRBP interacts with the distal end of Dicer's helicase branch, whereas Ago2 interacts with Dicer's catalytic branch. The structural information garnered in this study, combined with previous biochemical data, led to a testable model for how human RLCs efficiently transfer Dicer products to Ago2 in an orientation that allows for accurate selection of the targeting strand.

The preferred guide strand of any given siRNA or miRNA can be predicted with high accuracy based on the thermodynamics of the duplex ends. The RNA strand that has its 5´ end at the less stable end of the duplex is preferentially loaded onto Ago2 as the guide strand. In flies, an RLC subcomplex containing Dicer-2 and the dsRBP R2D2 is involved in sensing the thermodynamics of siRNAs prior to their loading onto Ago2, although it is unclear which protein is the actual sensor of thermodynamic asymmetry. In humans the sensor of siRNA thermodynamics has been unclear, given that Dicer forms a complex with either TRBP or PACT. In plants, some Dicer/dsRBP heterodimers are involved in sensing siRNA thermodynamics while others are not. Using in vitro biochemical experiments that were guided by structural insights, we show that human Dicer products are released and then rebound at a novel site along Dicer's helicase domain in a TRBP- or PACT-dependent manner. This novel binding site allows either Dicer/dsRBP heterodimer to sense siRNA thermodynamic asymmetry such that Dicer binds to the less stable end of the duplex and the associated dsRBP binds to the more stable end. We further demonstrate that Dicer itself is the sensor of siRNA thermodynamics and that this functionality is activated upon association with either TRBP or PACT. These crucial insights into siRNA positioning by human Dicer/dsRBP complexes led to a revised model for how strand-selective RISC loading may be achieved by RLCs in human RNAi pathways.

Although structural and biochemical data supports the notion that human RLCs function similarly to the Drosophila Ago2/Dicer-2/R2D2 RLC in strand-selective RISC loading, a number of clear differences exist. For example, in humans it has been shown that Ago2 can bind siRNAs and miRNAs in the absence of Dicer, which is not the case in flies. Furthermore, in certain cases Ago2 alone binds to duplex RNAs strand-selectively. The extent to which human RLCs are actually important for Ago2 loading and strand selection has therefore been controversial. We have used a reconstituted system to determine the degree to which each of the core components of the human RNAi machinery contributes to RISC loading and guide strand selection. We show that Ago2 has intrinsic but substrate-dependent strand selection capabilities. This activity, however, is in many cases enhanced substantially when Ago2 is in complex with Dicer and TRBP or PACT as a functional RLC. Our findings suggest that rather than functioning exclusively in human strand selection, Ago2's binding preferences serve instead as a secondary RISC loading checkpoint that acts in concert with Dicer/dsRBP asymmetry sensors to ensure proper strand selection by RLCs. The specific roles of each component in this process are dictated and fine-tuned by specific duplex parameters such as thermodynamics, 5´ nucleotide identity, and duplex structure. Surprisingly, our results also show that strand selection for some miRNAs is enhanced by PACT-containing complexes but not by those containing TRBP. Furthermore, overall mRNA targeting by miRNAs is disfavored for complexes containing TRBP but not PACT. This key finding reveals the possibility that RLCs containing TRBP may be optimized for the siRNA pathway, whereas RLCs containing PACT are optimized for the miRNA pathway. Overall, the body of work presented herein represents a significant step forward in our understanding of how core components of human RNAi pathways functionally interact within RLCs to achieve proper target specificity through strand-selective RISC loading.