42337 research outputs found
Sort by
Using future climate scenarios to support today’s decision making
Get PDFThis report reviews existing evidence and makes recommendations for the development of a practical climate scenario analysis tool to help public bodies adapt their infrastructure and service delivery for a changing climate
Measuring dark neutrinos and light using external and internal components of LArTPCs
No full textThe so-called MiniBooNE low-energy excess has been a long-standing question
for beyond Standard Model (BSM) physics in the neutrino field. The Short
Baseline Neutrino (SBN) program was proposed to investigate and provide further
insights into this excess. The SBN program utilises Liquid Argon Time Projection
Chamber (LArTPC) technology, which allows for precise measurements of neutrino
interactions and holds great potential for addressing significant open questions in
the field. Within the SBN program, the Short Baseline Near Detector (SBND) plays a
crucial role as one of the near detector, positioned just 110 meters from the neutrino
source. Being located on the surface, SBND is exposed to cosmic-ray background,
and to mitigate this, the Cosmic Ray Tagger (CRT) system has been implemented to
identify and veto cosmic background events, thereby enhancing the purity of neutrino
interactions. Additionally, the CRT system can be used to search for BSM signatures,
such as the dark neutrinos, which have been proposed to explain the MiniBooNE
low-energy excess. This thesis presents a novel approach for searching for dark
neutrinos using existing SBND CRT data, collected with a unique setup called the
CRT Beam Telescope. An extended model-independent search based on the dark
neutrino analysis is also presented. The sensitivity plot for the CRT Beam Telescope
is compared with an estimated MiniBooNE exclusion limit. Furthermore, this thesis
covers the commissioning work conducted for the SBND CRT system, which is vital
for ensuring optimal performance. The LArTPC technology also benefits from liquid
argon’s excellent scintillation properties, producing abundant scintillation light. This
thesis describes a method developed to understand the detector response to the
scintillation light produced across the LArTPC, enhancing our understanding of the
detector’s performance and facilitating more accurate measurements
Fair representation in participatory democracy
Get PDFThe goal of this thesis is to develop theoretical and empirical frameworks to promote fair and representative collective decision-making in Citizens’ Assemblies (CAs). In a CA, facilitators gather a demographically representative sample of the population. Participants listen to expert information on a specific topic and deliberate among themselves based on factors including their expert advice and their own lived experiences. They then generate a series of policy recommendations for the facilitating body. Empirical CA trials consistently demonstrate their ability to change participants’ beliefs, reducing any extremist views and helping them to reach consensus.
This thesis addresses four research questions to understand and augment the CA process using computational techniques:
1. How should algorithms support fair demographic representation and constructive deliberation? How can we optimise the group allocation algorithm to maximise pairwise interactions and maintain demographic representation from the wider panel, in a computationally unlimited setting?
2. How can algorithms support fair demographic representation and constructive deliberation in realistic settings? How can we optimise the algorithm used by facilitators while maintaining utility in terms of runtime and interpretability?
3. Why does better demographic representation in CAs lead to open-mindedness and affect deliberative outcomes?
4. When do CAs produce constructive deliberation, and what happens to the quality of debate when representation criteria are not met? Under what circumstances does fair demographic representation allow for constructive deliberation?
We develop two algorithmic approaches to support demographic representation: first in a controlled setting with computational and time resources, and then using a heuristic approach that performs well on past CA instances. We then construct an opinion model, calibrated to real data from the CA of Scotland, to explore the dynamics of the CA process and the importance of demographic representation in achieving an open-minded set of participants, which allows for engagement with experts and real opinion change. We finally demonstrate conditions that lead to successful outcomes using examples from 22 deliberative experiments, showing the real-world impact of achieving demographic representation.
Taken together, these works contribute important theoretical support for future CAs, as well as an algorithm to help optimise the process in future implementations
Kinetoplast DNA dynamics in trypanosoma species: the impact of life cycle variations and reproduction strategies
No full textTrypanosomatids are unicellular, flagellated obligatory protozoa parasites. Many dixenous trypanosomatids, such as trypanosome parasites in the genus Trypanosoma, cause diseases in humans and livestock. Human diseases due to trypanosome parasites mainly occur in developing or undeveloped countries, including Chagas in South and Central America (Trypanosoma cruzi), chronic Human African Trypanosomiasis (HAT) in Central and West Africa (Trypanosoma. brucei. gambiense type 1), and acute HAT in East Africa (Trypanosoma brucei rhodesiense). Meanwhile, Trypanosoma brucei brucei, Trypanosoma congolense, Trypanosoma brucei equiperdum, and Trypanosoma evansi afflict animals and cause Animal African Trypanosomiasis (AAT), nagana, dourine, and surra respectively.
The single mitochondrion of trypanosomatids contains a massive genome, the kinetoplast. Within an individual parasite, the kinetoplast DNA (kDNA) forms a chainmail-like network with two types of catenated DNA molecules: 20 to 50 copies of essentially identical maxicircles and thousands of highly heterogeneous minicircles. Maxicircle encodes ribosomal and electron transport chain subunits. The pre-mRNAs of 12 genes require post-transcriptional editing directed by short “guide RNAs” (gRNAs) encoded on minicircles. To produce translatable mRNAs, trypanosomatids must cover all editing sites with at least one gRNA. In species with extensive editing such as T. brucei, the kDNA network contains a highly diverse population of minicircles and encodes hundreds of distinct gRNAs.
The lifecycle of dixenous trypanosomatids involves insect vectors and mammalian hosts. During clonal reproduction, imperfect replication and segregation of kDNA may cause some minicircles encoding essential genes to drift towards a dangerously low abundance. In trypanosome parasites, sexual reproduction occurs exclusively in the insect vectors and results in mixing of the mitochondrial genome in the progeny. Circulating minicircles among tsetse-transmissible isolates, sexual reproduction potentially rescues low-abundance gRNA genes in the progeny by replenishing it with copies from the other parent. In addition, the different metabolisms at the mammalian and insect stages entail a different set of essential maxicircle genes and a lower demand for editing capacity in the mammalian stage. Consequently, deviations from the typical lifecycle can present a unique challenge in maintaining the kDNA integrity.
We propose that sexual reproduction is key in combating genetic diversity loss in kDNA. Conversely, the absence of sexual recombination reduces kDNA complexity. Using next-generation sequencing data, we have assembled and examined the kDNA from trypanosome species and subspecies with different life histories. Also transmitted by tsetse, the human pathogen T. b. gambiense type 1 reproduces strictly clonally. The clonal T. b. equiperdum and T. b. evansi no longer rely on tsetse but are transmitted directly between mammals. The dixenous T. congolense is colocalized with T. brucei and reproduces sexually in the proboscis of tsetse flies.
We report a highly conserved minicircle population characteristic of T. b. gambiense type 1 in 117 isolates. Comparing T. b. gambiense type 1 to the sexual T. brucei subspecies, we observed substantial kDNA streamlining in the asexual isolates with evidence of approaching tsetse transmissibility loss.
We confirm that in three groups of asexual kDNA-independent T. b. evansi and T. b. equiperdum, the minicircle genomes consist of thousands of a single minicircle class specific to each group. A putatively kDNA-independent ecotype, T. b. equiperdum group OVI retains moderate kDNA complexity and can probably produce fully edited mRNAs of A6 and RPS12, the only edited maxicircle genes required in mammalian bloodstream.
Comparison between T. brucei and T. congolense gRNA annotation revealed highly conserved editing blocks that cover the edited mRNAs with minimal overlaps. The results shed light on the evolution of the editing cascade
Optimal control theoretic value function learning
No full textGenerating behaviours to complete complex tasks can be viewed under the paradigm of controlling dynamical systems. To solve such tasks, most approaches fall under two paradigms: Reinforcement Learning (RL) and Optimal Control (OC) theoretic approaches. OC theoretic solutions are mostly local and can only provide global controllers for special cases. As a result of this, local solutions become trajectories. Synthesising these trajectories in the deterministic setting is formalised under the calculus of variations. This paradigm imposes strict constraints on the objective landscape: differentiability and continuity. In the stochastic setting, OC theoretic solutions have been proposed to remove the burden of these constraints and infer the optimal trajectory through sampling. RL has very similar theoretical groundings but diverges significantly in its approach. For example, RL parametersises value and/or policy functions instead of trajectories, allowing generalisation to new initial conditions. Additionally, in its model-free setting, which is our focus in this thesis, there is no need for constraints such as differentiability on the objective. RL is capable of estimating the gradients via sampling. However, these gradient estimates come at the high price of noisy solutions and slow convergence. To this end, defining methods that can leverage the best of both approaches is desirable. Our thesis aims to derive methods that greatly remove the burden of cost function design on the user while enabling generalisation by efficiently learning approximate global controllers.
As our initial attempt at this formalisation, we introduce a local method that combines the efficiency of derivative-based OC-theoretic approaches with the flexibility of local solutions based on sampling. To this end, we propose a hybrid approach that aims for consensus between the derivative-based solution of iterative Linear Quadratic Regulator (iLQR) and the sampling-based solution of Path Integral (PI) control. We define an objective that enables us to sample when derivatives vanish and follow optimised trajectories when derivatives arise. We use the Kullback Leibler (KL) control interpretation of PI control to formulate an inference problem that computes the optimal controls constrained by an adaptive distribution defined by the solution of iLQR. Our results show better convergence on manipulation and obstacle avoidance tasks than sampling strategy, path integral control and gradient-based strategy iLQR.
In the second segment of this thesis, we evaluate the widely used RL algorithms and its core gradient estimation machinery, policy gradients, without the typical convergence strategies. Our results are obtained on simple nonlinear continuous control problems. We show that RL still requires extensive tuning, even on simple nonlinear problems and the flexibility gained by zeroth-order derivative estimation is paid for by hyperparameter tuning. In turn, we propose an OC-theoretic approach based on Bellman optimality that leverages differentiable dynamics and first-order gradients. Our approach can learn approximate time-varying value functions and robustly converge with minimal tuning. We further verify the ability of our method by relaxing the objective and obtaining first-order approximations of time-varying Lyapunov constraints. We further verify our approach by satisfying this first-order constraint over a compact set of initial conditions. When comparing our method to Soft Actor-Critic (SAC) and Proximal Policy Optimisation (PPO) we show faster convergence and outperform PPO and SAC in task cost by at least 2 and 4 orders of magnitude, respectively.
In the third part of the thesis, we combine our findings from the previous sections to create a method that can handle discontinuities using stochasticity, ensure convergence with differentiability, and generalise with function parameterisation. To achieve this, we approach the problem using stochastic optimal control and robustness. We use the stochastic Hamilton-Jacobi-Bellman equation, differentiable dynamics, and the natural smoothing induced by stochastic first-order gradients. Our results demonstrate that the policies based on learned value functions outperform SAC and PPO in task cost by factors of up to 1076.02 and 8, respectively. Moreover, we observe that adding noise to the dynamics smoothens the curvature of the value function. This effect is especially noticeable in our obstacle navigation task with discontinuous dynamics and costs, where the value functions learned under noisier dynamics follow wider paths around obstacles, making them more robust. Finally, we show that our learned value functions can also be integrated into local methods, reducing their effective search horizon by a factor of 15
Seasonal and spatial dynamics of the diet and rumen microbiome of roe deer in Scotland
No full textRoe deer (Capreolus capreolus) are the most abundant and widespread wild ungulate species in Europe and play a significant role in shaping ecosystems through their browsing activities. Their impact on vegetation can lead to conflicts with forestry and landscape conservation objectives, which require effective management strategies. However, developing these methods requires a comprehensive understanding of the nutritional ecology of roe deer, which characterises how their diets and digestive processes respond to varying environmental conditions. Despite extensive research on roe deer ecology, significant knowledge gaps persist, particularly regarding the intricate relationships among their diet, local habitats, and gut microbiome across seasons and spatial scales.
This research addresses these knowledge gaps by providing a multifaceted investigation into the nutritional ecology of roe deer in Scotland. This research employs a combination of molecular techniques, including DNA metabarcoding, alongside remote sensing methodologies to characterise roe deer diets, rumen microbial communities, and local habitat composition across Scotland. By integrating these approaches, this thesis aims to elucidate how roe deer modify their foraging strategies and rumen microbiome in response to varying environmental conditions, with implications for both ecological understanding and management practices.
This thesis begins with a broad-scale biogeographical review of roe deer diets across Europe, setting the context for the subsequent focused study of Scottish populations. It then delves into a detailed analysis of the diet of roe deer across six sites in Scotland, using DNA metabarcoding of rumen contents to examine the composition and diversity of roe deer diets across different sites and seasons. The influence of local habitat characteristics on diet is then explored by integrating high-resolution remote sensing data with molecular diet analysis. The composition and diversity of the rumen microbiome are characterised, along with how it varies spatially and seasonally, and with intrinsic factors such as sex, age, and body weight.
Finally, the interactions between diet composition and the rumen microbiome are explored via both taxonomic and functional analyses. This multifaceted approach allows for an improved understanding of how roe deer modulate their foraging strategies and digestive processes in response to varying environmental conditions across Scotland. Key findings include the following: 1) Regional differences in the diet of roe deer across Europe. 2) There are significant seasonal and spatial variations in the diet composition and diversity of roe deer in Scotland, with summer diets being the most diverse and winter diets the least diverse. 3) There are strong associations between local habitat composition and diet, with certain habitat types predicting the presence and abundance of specific plant genera in the diet. 4) Seasonal shifts in rumen microbiome composition and diversity, with winter microbiomes being the most diverse. 5) An inverse relationship between dietary diversity and microbiome diversity, challenging previous assumptions about diet-microbiome interactions. 6) Individuals with similar diet compositions tend to harbor more similar rumen microbiome communities. 7) Complex, seasonally dynamic associations between specific plant genera and microbial taxa, suggesting interactions between the microbiome and dietary components.
This research advances the ecological understanding of roe deer and herbivore nutritional ecology more broadly, providing insights for wildlife management in Scotland and beyond. By exploring the complex relationships among habitat, diet, and the rumen microbiome, this thesis offers valuable insights into how wild ruminants respond to seasonal and spatial variations in resource availability. The integrated approach developed here, which combines DNA metabarcoding, remote sensing, and microbiome analysis, highlights the importance of considering both extrinsic environmental factors and intrinsic host characteristics in understanding wildlife nutrition. This multifaceted methodology offers a template for future studies, paving the way for more comprehensive approaches to wildlife ecology, conservation, and management
Muon neutrino and antineutrino cross sections & neutrino detector readout
Get PDFThe Micro-scale Booster Neutrino Experiment is a ground-level Liquid Argon
Time Projection Chamber neutrino detector, which collected data from the
Booster Neutrino Beam and coincidentally, the Neutrinos at the Main Injector
accelerator neutrino beam at Fermilab. The Deep Underground Neutrino
Experiment is a next generation neutrino detector experiment that aims to
provide answers to several critical questions posed by the current Standard
Model of particle physics. This thesis will explore the use of Machine Learning
techniques to extract interesting physics events from Liquid Argon Time Projection Chambers. It presents an investigation into the online triggering of data
readout, using a prototype DUNE far detector, based at CERN. An attempt
to separate an admixture of muon neutrinos and antineutrinos arriving at the
MicroBooNE detector from the NuMI flux is also presented. The analysis then
proceeds to extract charged current cross sections for the separated neutrino and
antineutrino charged current interactions on argon nucleons. An understanding
of these interactions is critical to a successful DUNE physics campaign
The difficulty of morality and the difficulty of moral psychology: some methodological and theoretical considerations
Get PDFThis thesis offers a novel critical examination of influential approaches and methodologies in empirical moral psychology by elucidating the hitherto overlooked convergence in critiques from psychology, philosophy of science and twentieth-century moral philosophy. The general aim is to show that there are some pervasive and sometimes peculiar methodological and theoretical issues pertaining to moral psychology. These issues prevent experimental findings from attaining a high degree of validity and/or reliability. At the heart of it all is the difficulty of studying a subject matter as complex as morality and the significant impact of researchers’ values and interests on research outputs.
Chapter 1 is the introductory chapter, where I spell out the context from which the relevant issues arise and provide an outline of the thesis. Chapters 2 to 4 constitute the negative critique, while Chapter 5 presents the positive proposal.
In Chapter 2, I discuss some issues with influential empirical approaches to study utilitarian thinking. The first issue concerns moral psychological studies that draw explicit connection between sacrificial dilemmas and normative ethical theories, such as utilitarianism. This approach reflects what I call an “assumption of fit with theory” – the assumption that sacrificial dilemmas reliably track attitudes towards, or inclinations about, certain ethical theories. I argue that there are good reasons to question this assumption. This casts doubt on the experimental findings that allegedly shed light on utilitarian thinking. To address the assumption of fit, some researchers propose paradigms that move beyond using sacrificial dilemmas as stimulus materials. I argue that when the proposed research paradigm involves studying a broad prescriptive recommendation, further problems arise.
In Chapter 3, I cast doubt on the ecological validity of moral psychological lab studies, through examining studies that use sacrificial dilemmas as stimulus materials. I first argue that the hypothetical sacrificial dilemmas used as stimuli in laboratories are often unrepresentative of real-world moral situations. I then argue that simply changing the stimuli does not resolve the issue. This is because there are some potentially important features of real-world moral behaviour and experience (e.g., relationality, subjectivity, context-dependency, creativity and unpredictability) that certain experimental methods (e.g., using stimulus materials in laboratories and controlling for confounding variables) cannot accommodate. However, even if one rejects this characterisation of moral behaviour and experience, there is still a cause for concern. This is because methodologies employed to design and conduct studies aim at experimental control. Experimental control comes in tandem with abstraction, which is characterised by the omission of details. The mere possibility that whatever lies outside of the confines of a lab study is in fact morally significant warrants concerns about ecological validity.
In Chapter 4, I discuss issues surrounding what I call “fragmentation” in moral psychology. Fragmentation involves breaking a target of inquiry into constitutive parts or categories, with the assumption that the identified parts are fundamental to the target, and invoking these parts explains the target’s behaviour. In moral psychology, fragmentation is seen in the way in which moral foundations are identified and isolated, and an explanation of the moral behaviour of interest is furnished by reference to the moral foundation(s) alone. It is also seen in the way in which a small number of categories of psychological processes, e.g., automatic emotional processes and controlled reasoning processes, are invoked as the fundamental processes underpinning deontological and utilitarian judgements. More broadly, fragmentation manifests in the way in which neurocognitive properties and processes are fragmented from other features relevant to a target moral phenomenon.
I demonstrate that worries about fragmentation raised elsewhere can be applied to contemporary theorists’ approaches in moral psychology. I highlight that firstly, fragmenting tendencies might be counterproductive to understanding a subject matter that is by nature complex, multifaceted and pluralistic, such as moral subjects or phenomena. I then argue that there is room to question the extent to which the morally relevant foundations, processes or mechanisms identified comprise stable separable parts, which display reliable behaviour across different contexts. Finally, I address two instrumentalist responses to my critique and argue that adopting an instrumentalist or pragmatist stance does not readily escape the danger of fragmentation.
In light of the issues discussed in Chapters 2 to 4, I gesture towards some recommendations in Chapter 5 for future research on morality in moral psychology. While some methodological issues may be resolved by correcting, changing or adopting new methods, some other issues that I have discussed are resistant to simple, methodologically oriented solutions. This is because solutions that merely target methodologies still operate within the confines of, and thus do not eliminate, the underlying theoretical issues. The aim of the positive proposal is in part to identify which of the issues discussed previously can be resolved, diminished or avoided in a minimally problematic way, and offer some considerations for how to achieve that. The general lesson is that a paradigmatic shift towards more hypothesis-generating and theory-building approaches can alleviate concerns about ecological validity significantly
Advancing human-computer interaction: acoustic-based intelligent hand tracking and recognition for the future
Get PDFAcoustic-based hand tracking technology is heralding a new era in Human-Computer Interaction (HCI) by utilising the speakers and microphones already embedded in many commercial devices. This approach leverages acoustic signals to detect and process echoes, thereby determining the position of a user’s hand in space.
This thesis commences with the construction of a circuit prototype designed to refine algorithm development and reduce the constraints associated with direct testing on
commercial devices. The prototype features adjustable placements for speakers and microphones, allowing for flexibility in their positioning. Initial evaluations on this
prototype demonstrated a 1D hand tracking accuracy of 3.7 mm, also confirming the system’s capability to accurately track 2D trajectories and visually recognise them.
A key application of acoustic-based hand tracking explored in this thesis is airwriting, where the user can write characters in mid-air, and the system records and recognises the resulting trajectories. Existing tracking approaches do not support
multi-stroke input, as a result, the trajectory is incapable of character recognition with models trained by open source hand written characters images. As a result, a virtual writing system is proposed to estimate the status of the hand with the energy information acquired from the echo. Subsequently, Zadoff-Chu (ZC) sequences and sinusoidal waves are used to obtain the initial position of the hand, and track the hand
continuously with the change of phase for a smooth trajectory. While inputting the characters, the virtual writing system allows the user to input multiple strokes to get rid of redundant trajectory which affects the recognition. The results show that the virtual writing system achieves an average error of 4.3 mm for 2D tracking and 94.8% recognition accuracy for 52 English letters, 10 numbers, and 20 Chinese characters.
Another critical challenge with existing approaches is the limited accuracy in accommodating multiple users, which restricts their utility in scenarios such as multiplayer gaming. A dual-user interaction system is then proposed to enable precise hand tracking for cursor control. Notably, it supports two users to operate independently. For the first user, a dedicated set of one speaker and two microphones are employed. To enable the participation of a second user, an additional speaker and two supplementary microphones were positioned in the opposite direction. To minimise the interference between users, distinct frequencies of sinusoidal waves and ZC signals generated by the co-prime root indexes were allocated for each user. The proposed method demonstrates remarkable performance, achieving an average
hand tracking error of 5.7 mm, and maintains consistent accuracy in dual-user scenarios.
Finally, the thesis explores the expansion from 2D to 3D hand tracking, a necessary evolution to capture more complex hand gestures and movements. Implementing a 3D tracking system requires at least three, and ideally four microphones to accurately compute hand coordinates in 3D space. Modifications to the original prototype allow for the integration of additional microphones, and a robust mathematical framework is developed to translate multiple distance measurements into precise 3D coordinates,
achieving an average 3D hand tracking error of 1.9 cm with four microphones.
In summary, this thesis advances the field of HCI by developing robust and innovative solutions for acoustic-based hand sensing, specifically designed for close-range interactions within 5 – 30 cm. It lays the groundwork for more natural and intuitive user interactions, providing a strong foundation for future research aimed at refining these systems and integrating them into everyday technological applications
Comparative genomics of mobulid rays using a novel West Atlantic pygmy devil ray (Mobula hypostoma) reference genome
Get PDFElasmobranchs are a subclass of cartilaginous fishes comprising around 1200 species of sharks, rays and skates. A combination of low-reproductive history traits and threats such as habitat destruction, overfishing and climate change, has led to a dramatic decline in elasmobranch population numbers. In recent years, elasmobranch conservation and stock management have increasingly relied on molecular data to understand speciation, habitat use and population dynamics; simultaneously, the past decade has seen an increased recognition of elasmobranchs as an important deep-branching clade in comparative studies aimed at understanding vertebrate evolution. Therefore, a growing body of high-quality, whole-genome sequences have been assembled for several elasmobranch species across most orders, including mobulid rays.
Mobulid rays, manta and devil ray species of genus Mobula, are highly derived batoid fishes with unique morphological features including cephalic lobes framing a forward-facing mouth; in contrast to ancestral ray and skate lineages, who are typically adapted to benthic environments, mobulids have evolved to expand into pelagic niches. Despite their endangered status, mobulid rays remain poorly studied and understood among cartilaginous fishes, including the genomic basis for their ecological adaptations and possible susceptibilities to environmental change. Therefore, the assembly of a novel chromosome-level, high-quality reference genome for the Atlantic pygmy devil ray, Mobula hypostoma, provides a unique opportunity to investigate mobulid evolutionary history and adaptation, and support future conservation studies.
Whole-genome alignment of M. hypostoma, with its sister species oceanic manta ray (M. birostris), Atlantic stingray (Hypanus sabinus) and two skate (Leucoraja erinacea and Amblyraja radiata) reference assemblies, was carried out using Cactus. De novo repeat libraries were constructed using RepeatMasker to characterise and compare the repeat landscape of each aligned species. Synteny analysis, homology relationships and structural rearrangements between these genomes were investigated to highlight mobulid-specific mutations. Functional annotation of mobulid – and M. hypostoma – specific genomic rearrangements was carried out to investigate their effects on genes and phenotype.
Results showed one-to one chromosome homology was mostly maintained between the two mobulid species, reflecting their recent divergence; in contrast, more substantial rearrangements were observed between mobulids and stingray, mostly chromosomal fusions. Repeats accounted for most of the genome content for all species (60-65%), with M. hypostoma possessing the highest genome content across all five batoids, and the highest percentage of DNA transposons and retroelements. Genomic regions lost in the mobulid ancestor directly impacted or were located near to many candidate genes associated with vertebrate craniofacial development, including alx3, fgf4, foxe1 among others. Further research will be needed to determine if these genomic changes were responsible
Access Repository Dashboard
Do you manage Edinburgh Research Archive? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!