11319 research outputs found
Sort by
Is the Pen Mightier than the Sword? Testing the Effects of Fiction on Empathy and Aggression, and Introducing a Hybrid Model of Media Effects
No full textExposure to violent television and video games has been linked with increased aggression and decreased prosocial behaviour. However, little is known about whether these effects appear for violent literature. Conversely, reading fiction has been linked to increased social cognition, but it is unclear whether violent literature content would also show this benefit. This thesis aims to integrate these two fields by extending the framework of the General Aggression Model to include fictional transportation and its interaction with media content. Further, these studies investigate whether reading violent literature increases aggression and decreases pro-social behaviour in a sample of emerging adults, compared to watching violent television. Three experimental studies compared the effects of watching to reading violent content on aggression and prosocial behaviour. It was hypothesised that violent television and literature would increase aggression and decrease prosocial behaviour, but the effects for reading would be weaker than those for television. The first experiment yielded null results of media condition on all outcome measures but did have several limitations. The second experiment found that watching and reading violence decreased prosocial behaviour compared to a control. Reading violent fiction also increased social cognition scores compared to the other two conditions. The third experiment demonstrated that reading violent content increased aggression, but these effects were moderated by trait aggression and empathy, partially supporting the hypotheses. Two correlational studies then investigated the long-term links between watching and reading fiction and empathy and aggression. It was predicted that greater fiction-watching would show similar but weaker associations than fiction-reading in terms of higher empathy and lower aggression. However, only fiction-reading preferences were linked to increases in empathy and decreased aggression, partially supporting the hypotheses. A new Hybrid Model of Media Effects is presented to explain these findings
Detection of Adulteration in New Zealand High Value Dairy Products Using NMR-Based Metabolomics: A Chemometrics Approach
Get PDFNew Zealand (NZ) is the world’s largest exporter of dairy products. NZ milk products are known for their ‘clean and green’ reputation with high nutritional properties. In fact, the quality of NZ milk is perceived to be one of the best milk in the world. Cow, goat, and sheep milk are produced in NZ. In fact, NZ goat milk (GM) and sheep milk (SM) are considered as high-value products. The price of goat and sheep milk is 2-3x the price of cow milk (CM) in NZ. These facts put these high-value dairy products at risks of fraudulent activities like milk adulteration or counterfeit. Milk can be adulterated with different adulterants including water, vegetable protein, whey, and milk from another species. In other cases, expensive milk (e.g., GM, SM) could be adulterated with cheap milk (CM) for the purpose of economic gain. This is also known as economically motivated adulteration (EMA). It is therefore important to develop quick, effective, and robust tools for the detection of adulteration. Such techniques must be robust and high throughput requires a small amount of sample, and highly reproducible. Of particular interest is the application of proton nuclear magnetic resonance (1H-NMR) fingerprinting technique, which meets most of the mentioned requirements.
To date, there is limited information on the metabolite composition of milk produced in NZ. Moreover, NZ milk are not well studied in terms of their compositional properties. Therefore, this thesis is aimed to explore the capability for NMR-based metabolomics technique in detection of adulteration of NZ GM and SM with different concentrations of CM. To achieve this, the study was split into two parts. In the first part, NMR spectroscopy was used to characterise NZ CM, GM, and SM powder to select the discriminant metabolites for each species. In the second part, NMR was used to detect adulteration of GM and SM with different concentrations of CM. Advanced chemometrics (supervised and unsupervised approaches) were applied for data interpretation. SPSS and R studio were also employed for statistical analysis.
Overall, NMR fingerprinting technique alongside advanced chemometrics enabled detection of 17, 24, and 23 metabolites present in the water-soluble fractions of CM, GM, and SM, respectively. Out of the identified metabolites, carbohydrates, carboxylic acid, and amino acid were amongst the selected discriminant compounds in CM. In GM, the selected discriminant compounds include amino acid, fatty acid, nucleosides, carbohydrates, and carboxylic acid. Lastly, compounds such as carboxylic acid, carbohydrate, and nucleotide were selected as discriminant markers of SM. Following characterization, NMR spectroscopy was also successful in identifying potential markers of adulteration in GM and SM with CM. Based on VID feature selection procedure and Tukey’s test, phosphocholine was selected as a candidate marker of adulteration of GM with CM. On the other hand, N-acetyl carbohydrates and orotate can be proposed as potential markers of adulteration of SM with CM.
This work is the first study to characterize NZ milk types (CM, GM, SM) using NMR-based metabolomics, and attempt to detect adulteration of NZ GM and SM with different concentrations of CM. Overall, the NMR-fingerprinting technique was successful in characterising the metabolites present in the different milk types and detecting adulterations. Advanced chemometrics (supervised and unsupervised approach) were also suitable for the interpretation of NMR data, and for identifying discriminants, and in detecting adulterants. Further investigation of different milk fractions (such as lipid) and also the use of other fingerprinting techniques (e.g., LC-Q-TOF-MS, infrared) is needed to support the findings of the present study.
Keywords: NMR, metabolomics, chemometrics, cow milk, goat milk, sheep milk, adulteration
detection, metabolites, marker
Localisation for Augmented Reality at Sports Events
No full textThe use of computer generated graphics and visualisations has become increasingly common in professional sports broadcasting over the past two decades. Overlaid graphics are designed to convey information about the game to the spectator in a way that is contextualised spatially, such as virtual offside lines being overlaid on soccer footage.
We aim to take these graphics, that are only seen in professional broadcasting, and provide them to on-site sports spectators using an augmented reality mobile application. Allowing fans in the stadium to view the same pitch-aligned visualisations as those watching at home, all rendered on their phone in real time.
In sports broadcasting the alignment of these graphics with the environment is often achieved by using professional cameras equipped with high accuracy sensors or elaborate manual calibration techniques to measure the broadcasting camera's position and orientation. These measurements allow the graphics to be accurately matched to the camera view.
The goal of our research was to design a method of camera localisation that will allow spectators to enjoy accurately aligned augmented reality content with minimal prior calibration and without a complicated manual initialisation process. We investigated whether line pitch markings could be used to estimate an on-site user's position and orientation with sufficient accuracy to align augmented reality visualisations with the pitch. Our evaluation found that our approach performed with sufficient speed and accuracy to form the basis of a mobile augmented reality application for on-site spectators
Kisspeptin regulation of oxytocin neuron activity in late pregnancy
No full textOxytocin triggers uterine contractions during parturition. My laboratory recently showed that the oxytocin neurons are excited by intracerebroventricular (ICV) administration of kisspeptin only in late-pregnant rats. Furthermore, a kisspeptin projection from the periventricular nucleus (PeN) of the hypothalamus to the perinuclear zone (PNZ) surrounding oxytocin neurons in the supraoptic nucleus (SON) emerges in late pregnancy. The research described in this thesis aimed to determine the site and mechanism of central kisspeptin activation of oxytocin neuron activity in late-pregnant rats.
To determine whether ICV kisspeptin directly activates oxytocin neurons, I performed immunohistochemistry for the kisspeptin-activated second messengers, phosphorylated extracellular regulated kinase 1/2 (pERK1/2) and phosphorylated-p38 (p-p38). I found that ICV kisspeptin (2 µg in 1 µg/µl) did not change pERK1/2 or p-p38 expression in oxytocin neurons in the SON and PVN, suggesting that exogenous kisspeptin does not directly activate oxytocin neurons.
I then investigated pERK1/2 and p-p38 expression in brain areas that project to oxytocin neurons to determine whether kisspeptin acts indirectly at cell bodies of afferent inputs to excite oxytocin neurons. ICV kisspeptin did not change pERK1/2 or p-p38 expression in the nucleus tractus solitarius (NTS), rostral or caudal ventrolateral medulla (VLM), or dorsal raphe nucleus (DRN) in the brainstem, or in the organum vasculosum lamina terminalis (OVLT), median preoptic nucleus (MnPO), subfornical organ (SFO), or anteroventral periventricular nucleus (AVPe)/PeN in the forebrain. Hence, exogenous kisspeptin does not appear to activate afferent inputs to oxytocin neurons.
To determine whether kisspeptin acts locally within the SON, in vivo extracellular single-unit recordings were made from SON neurons during intra-SON microdialysis administration of kisspeptin (100 µM in the dialysate) in urethane-anaesthetised rats. I found that intra-SON kisspeptin consistently increased the firing rate of oxytocin neurons in urethane-anaesthetised late-pregnant rats but had no effect in non-pregnant rats. Hence, it appears that kisspeptin excites SON oxytocin neurons by a local action within the SON.
Noradrenergic projections to the SON from the NTS are prominently activated at parturition. Therefore, I next investigated whether kisspeptin enhances noradrenergic signalling to oxytocin neurons in late pregnancy. To determine whether there is an interaction between kisspeptin and TH-positive (a marker for noradrenaline neurons) fibres in the SON/PNZ, I performed immunohistochemistry for kisspeptin and TH in the SON/PNZ. Confocal image analysis showed similar low levels of colocalisation/appositions between kisspeptin and TH in the SON/PNZ of non-pregnant and late-pregnant rats, suggesting there is no increase in putative synapses between kisspeptin and TH fibres in late pregnancy. Next, I investigated whether kisspeptin pre-synaptically modulated excitatory noradrenergic inputs to oxytocin neurons in late pregnancy using in vivo electrophysiology to determine the functional interaction between kisspeptin and noradrenergic inputs to oxytocin neurons. Intravenous (IV) cholecystokinin-8S (CCK-8S) excites oxytocin neurons via noradrenergic inputs. Hence, the oxytocin neuron response to IV CCK-8S was determined before and during intra-SON administration of kisspeptin. Preliminary analysis showed no significant differences in oxytocin neuron responses to IV CCK-8S before or during kisspeptin administration in non-pregnant and late-pregnant rats, suggesting that kisspeptin excitation of oxytocin neurons might not be mediated by noradrenergic inputs.
Taken together, these results suggest that kisspeptin might excite oxytocin neurons at the end of pregnancy by a presynaptic action at their afferent inputs, but the specific input involved is yet to be identified
Food Waste in the Cafe & Restaurant: Sector in New Zealand
Get PDFIn 2017/2018, research was conducted into the food waste generated by restaurants and cafes in New Zealand. The research found that 24,375 tonnes of food waste is generated per annum. Of this food waste 61% is avoidable and 39% is unavoidable e.g. eggshells, banana skins, bones etc. The average cafe and restaurant generates approximately 2.8 tonnes of food waste per annum.
When cafes with a similar menu and offering were compared, the difference in the amount of avoidable food waste between the best and worst performing cafes was 43%, showing that there are clear opportunities for some cafes to reduce their food waste.
The aim of this research was to quantify the amount of food wasted by cafes and restaurants in New Zealand.
In August-September 2017, a master’s student from the University of Otago conducted nine bin audits at five restaurants and four cafes at six different locations around New Zealand.
In April 2018, WasteMINZ employed an intern to conduct audits of restaurants and cafes in Auckland. In total 10 cafes and one restaurant were audited
Comparative and functional genomics of the Lotus microsymbiont Mesorhizobium japonicum R7A
No full textSymbiotic mesorhizobia are soil bacteria which have coevolved with diverse integrative and conjugative elements that confer the capacity for nitrogen-fixing symbiosis with legumes (ICESym). The isolate Mesorhizobium japonicum R7A is a model mesorhizobial symbiont of plants in the Loteae tribe of legumes, including the model plant Lotus japonicus Gifu. Wildtype M. japonicum R7A (R7AWT) carries the symbiosis island ICEMlSymR7A, and a derivative strain cured of the ICESym, M. japonicum R7ANS (R7ANS), has previously been developed. Through acquisition by horizontal transfer of different ICESyms, R7ANS is conferred a symbiotic host-range dependent on the specific ICESym it harbours.
The rhizosphere is the zone of plant influence which surrounds the roots of a plant in the soil. The rhizosphere effect is the change in the rhizosphere affected by a growing plant on the soil and the microbiota therein. Research suggests that the rhizosphere effect of plants is a directed process that may actively shape microbial communities. Little is known about the growth of R7AWT or R7ANS in the rhizosphere of L. japonicus or Arabidopsis thaliana, or what effect the ICEMlSymR7A has on competition therein. Here I present comparative and functional genomic research on the basis of ICESym host range and R7A rhizosphere competition.
Through comparative genomics of Loteae nodulating ICESyms, I identified an evolutionary relationship between them that is correlated with their conferred host ranges. I identified that across the Loteae nodulating ICESyms compared, 155 genes were conserved representing on average 30% of a given ICESym’s genes. I also found that on average 35% of an ICESym’s genes were unique to it. A total of 53 genes and 4 nod box symbiotic regulatory motifs were uniquely conserved in the broad host range L3 lineage of Loteae-ICESyms, and I present a discussion of these genes and how they may account for the differing symbiotic host ranges.
Using the functional genomic screening method Tn-seq, I screened over 800,000 unique Tn5 mutants in both R7AWT and R7ANS, identifying 600 core essential genes for growth in vitro while also identifying that ICEMlSymR7A modulates the essential phenotype of the house-keeping genes pncB, groELS and a lysine tRNA gene. I also identified 19 genes with essential phenotypes encoded by ICEMlSymR7A which may be the result of toxic effects caused by Tn insertion. Some of these genes resembled defence systems to mobile genetic elements. The Tn-seq results indicated the Entner-Doudoroff pathway was essential for catabolism of glucose, and this was confirmed by insertional mutagenesis.
Finally, I used Tn-seq to screen symbiotic and non-symbiotic rhizosphere competition of R7AWT and R7ANS identifying 129 pan-rhizosphere genes in L. japonicus and A. thaliana rhizospheres that resulted in decreased competitiveness when mutated. In 1:1 competition assays I observed that ICEMlSymR7A conferred an increase in rhizosphere competitiveness, but that Nod factor signalling conferred a relative fitness cost. Analysis of metabolic genes with a loss of fitness in the rhizosphere Tn-seq screens indicated that metabolism of organics acids was important for rhizosphere competitiveness, which may relate to the decreased competitiveness observed for R7ANS.
Mesothelioma by Accident
No full textPresents an overview of whether cover is available in New Zealand for those who have developed mesothelioma as a result of an accident. Cites the recent High Court decision 'Calver v ACC' [2019] 3 NZLR 261 [2019] NZHC 1581, which will go to the Court of Appeal ('ACC v Calver' [2019] NZHC 2667). Describes the diagnosis of mesothelioma and the complex provisions for cover in the Accident Compensation Act 2001. Focuses on the provisions in this Act regarding whether there is a personal injury and the circumstances in which the injury was sustained. Explains that three different answers are possible in classifying mesothelioma as a personal injury, and details and comments on these situations. Discusses the vexed question that Courts need to consider to determine whether mesothelioma was caused as an inhalation accident, and cites relevant cases.Peer Reviewe
Designing isothiocyanates that target macrophage migration inhibitory factor
No full textIsothiocyanates (ITCs) are a class of compounds found in edible cruciferous vegetables. They have been extensively studied both in vitro and in vivo for their preventive and therapeutic effects in number of diseases, including inflammation and cancer. Increasing epidemiological evidence demonstrates that increased consumption of cruciferous vegetables leads to improved outcomes in many cancers and inflammatory diseases. These benefits may be attributable to ITCs. Evidence suggests that ITCs exert their biological effects through distinct but interconnected signalling pathways.
Macrophage migration inhibitory factor (MIF) is a proinflammatory cytokine with well-characterised roles in inflammation, and innate and adaptive human immune responses. MIF is overexpressed in various cancers, including colorectal cancer, pancreatic cancer, leukemia, prostate cancer and breast cancer as well as inflammatory conditions such as acute pancreatitis, septic shock, rheumatoid arthritis and atherosclerosis. MIF exerts its biological activity by binding to the cell surface receptor CD74 as well as intracellular signalling proteins. MIF also possesses keto-enol tautomerase activity against dopachrome, dopachrome methyl ester, phenylpyruvate and p-hydroxyphenylpyruvate with an N-terminal proline as the key catalytic amino acid. Whilst no physiological substrates of MIF have yet to be identified, inhibition of tautomerase activity has been associated with a loss of biological function of MIF.
Recent studies have identified small molecule MIF inhibitors for treating inflammatory diseases and other pathologies. In 2009, our research group identified ITCs as irreversible inhibitors of MIF’s tautomerase activity via covalent attachment to the N-terminal catalytic proline. Benzyl and phenethyl ITC derivatives that orient in “upward” binding mode have become attractive lead compounds for development of MIF inhibitors, especially with a sulfonamide substituent.
In this thesis, a library of ITCs featuring benzyl, phenethyl and phenyl propyl isothiocyanates with sulfonamide and carboxylate substituents were designed, synthesised and purified to determine whether the acidic nature of these substituents or hydrogen bonding characteristics are essential for inhibiting MIF. The derivatives were expected to bind in “upward” orientation. Ten ITCs synthesised from commercially available starting materials were tested for their ability to inhibit recombinant human MIF (rhMIF) and MIF activity in Jurkat T-lymphoma cells. rhMIF was expressed in Escherichia coli and purified using three protocols in an attempt to obtain a homogenous form with post-translational cleavage of N-terminal methionine. All the three protocols produced heterogeneous form of protein with some of the protein retaining the N-terminal methionine. Among the three, protein obtained from a single preparation (protocol 1) was used to assess the synthesised ITCs as it contained the least amount of inactive protein.
ITCs with sulfonamide substituents LPR07, LPR16, LPR17, LPR25 and LPR26 showed good MIF tautomerase inhibition with submicromolar IC50 values in both enzyme and cell-based assays. ITCs with carboxylate substituents LPR06 and LPR08 showed reduced MIF tautomerase inhibition. Neither showed strong inhibition in a cell-based assay. The phenethyl (LPR25) and phenyl propyl (LPR26) derivatives of racemic sulfonyl methylpiperidine ITC (LPR17) were synthesised and assessed. No correlation was observed between activity with differing linker length (from one to three carbons). Both enantiomers of LPR17 were synthesised to identify the more active stereoisomer. The R enantiomer (R-LPR29) was shown to be more potent than the S enantiomer (S-LPR32) in both rhMIF and Jurkat cell MIF tautomerase assays. LPR16 and LPR17 showed good inhibitory activity in both assays. The binding affinity, Ki, rate of inactivation, kinact and specific reactivity, kinact/ Ki of PEITC, LPR16 and LPR17 was studied using a two-step covalent inhibition mechanism under special experimental conditions and analysed using DynaFit software. LPR17 showed low reactivity with high binding affinity and LPR16 showed high reactivity with less binding affinity. However, a linear correlation between the specific reactivity and IC50 values of LPR16 and LPR17 was not observed.
The docking studies of LPR16, LPR02, LPR07, R-LPR29, S-LPR32, LPR25 and LPR26 showed cis and trans thiourea bond formation with “upward” binding orientation as expected and the potential for sulfonamide hydrogen bonding and π stacking interactions with the adjacent Tyr-37 favoured potent MIF tautomerase inhibitory activity. The hydrogen bonding characteristics of the sulfonamide ITC derivatives are required for MIF inhibition but not the acidic nature of either sulfonamide or carboxylate substituents.
The lead molecules LPR16 and LPR17 were examined for their cytotoxicity against CT26 colorectal cancer cells. LPR16 and LPR17 showed dose-dependent cytotoxic potential towards CT26 colorectal cancer cell viability. LPR16 and LPR17 were also evaluated by the National Cancer Institute (NCI) 60 human tumour cell lines screen (NCI60) at 10 µM. The single-dose response data of LPR16 and LPR17 showed good growth inhibition of many cancer cell lines, especially gastrointestinal cancer cell lines. At 10 µM, LPR16 and LPR17 reduced the growth (below 60% growth percent) of colon (COLO 205, HCT-116, HCT-15, HT29, KM12 and SW-620) and renal cancer cells (ACHN, CAKI-1 and RXF 393). LPR16 was selected for further screening and evaluated at five doses (10 nM to 100 µM). LPR16 showed strong growth inhibition and cytotoxicity against colon cancer, renal cancer, leukemia, melanoma, ovarian and breast cancer cell lines.
In summary, these results show that MIF’s tautomerase activity was inhibited by synthesised ITCs with the majority more potent than other reported ITCs and other MIF covalent inhibitors. There is a considerable scope for further development of ITCs as potent MIF tautomerase inhibitors for different diseases including gastrointestinal cancers or inflammatory diseases
Error decomposition of evolutionary machine learning
Get PDFAlgorithms or models are often measured using a fitness function that calculates total prediction error. While reducing total error is typically the overall objective, examining error as an aggregate value does not provide a thorough understanding of model behaviour. A greater appreciation of behaviour can be obtained through performing a bias-variance decomposition to split error into bias and variance components. This is effective for assessing a deterministic algorithm, such as Classification and Regression Trees (CART). However, splitting the error into bias and variance is not sufficient for non-deterministic algorithms, such as genetic programming (GP), that potentially produce a different model each time they are applied to the same sample.
This thesis presents an extended bias-variance decomposition that decomposes error into bias, external variance (error attributable to limited sampling of the problem), and internal variance (error due to random actions performed as part of the algorithm’s execution). Using this extended decomposition framework provides a greater understanding of the behaviour of an algorithm, and enables more informed choices in algorithm refinement to be enacted. While the extended error decomposition is primarily applied to GP, it can be used to improve the understanding of any non-deterministic algorithm. This is important for interpreting the modelling properties of these types of algorithms, irrespective of their complexity or assumptions regarding their behaviour.
In this thesis, the extended decomposition is used to characterise the behaviour of GP variants in the literature, determining whether they behave as reported in terms of reducing a particular component of error. While the majority of this thesis examines synthetic data sets, methods for decomposition using finite data samples are also examined to determine their suitability for error analysis.
The application of Z-score standardisation (zero mean with unit variance) to GP is rare in the literature. In this thesis, the extended decomposition is applied to a variant of GP using standardisation. It is shown to generally improve the predictive performance of GP by reducing error due to bias. However, it can also be associated with erratic error due to variance (particularly internal variance). This thesis uncovers evidence that this is due to sparse examples of the training data near the boundary intervals of the explanatory variable. A simple solution to this problem is presented, which involves augmenting the training data with observations at the boundaries of these intervals. Cross-validation is used to determine the appropriate number of boundary observations for a number of “real-world” data sets. The results show that standardisation generally improves the predictive performance of GP for these data sets, often without the need for augmentation, demonstrating the importance and reliability of applying standardisation to GP.
This thesis proposes initial steps towards determining how the extended decomposition can be used for algorithm refinement by targeting a reduction of the largest component of error. This is examined for automated algorithm refinement by comparing different combinations of algorithm modules. If these modules have been selected as candidates after characterising them solely using total error (or without characterising them), they may not provide a sufficiently diverse range of behaviours for comparing and combining them effectively. Also, this thesis demonstrates the difficulty in comparing algorithm modules with relatively similar decomposed error when error due to internal variance is unstable. Alternatively, algorithm adaptations with well-understood behaviour in terms of decomposed error are examined for manual algorithm refinement. This highlights the need for automatic machine learning methods that consider a diverse portfolio of modules in terms of targeting the reduction of different components of error for a given problem
Loneliness and wellbeing amongst New Zealand Veterans
No full textObjective: To identify the proportion of New Zealand (NZ) veterans who suffer from loneliness, and whether this is correlated with general wellbeing. In addition, we wished to find out whether NZ veterans are aware of supports for loneliness and feel comfortable accessing these supports.
Methods: Current and previous New Zealand veterans aged at least 18 years of age who are members of Royal NZ Returned and Services Association (NZRSA) and/or follow NZ military Facebook sites were provided with a study information sheet and a URL for an online survey. The survey included questions on demographics, loneliness (in the form of a one-item loneliness question and the UCLA-3 three-item loneliness scale), wellbeing (in the form of the General Health Questionnaire-12) and a question indicating which existing support agencies they were aware of.
Results: The proportion of loneliness in the sample was 33.0% (95% CI 23.8%-43.6%). A correlation was found between loneliness and psychological distress, indicating that as loneliness increased psychological distress appeared to also increase (Pearson correlation 0.63). 14.8% of all respondents were not aware of any of the supports listed. A higher percentage, 21.4%, of those respondents who were lonely were not aware of any of the supports. 43.2% of all respondents would not feel comfortable accessing supports and this was also greater for those who self-reported as lonely (55.2%).
Conclusions: The findings in this study showed that loneliness is likely a widespread problem amongst NZ veterans. This is an area which warrants further research, which might aid in identifying potential supports to minimise the impact of loneliness amongst the NZ veterans