Exposure to traumatic events during infancy can lead to adverse health effects later in life, stemming from an imbalance between the innate and acquired arms of the immune system. Trauma experienced at an early age, such as childhood abuse or environmental stress, is categorized as early life adversity. As the immune consequences of early life adversity are unclear, our study explores how early life adversity affects immunological development and parasite susceptibility in gelada monkeys (Theropithecus gelada). Geladas are an ideal species to study early life adversity due to the similarity of their immune response to that of humans and the fact that they may face adversity early in life. Male geladas compete for reproductive access to females, which often leads to attempted infanticide. Surviving juveniles experience trauma that may have similar physiological consequences to early life adversity in humans. Thus, we determine if there are differences in the immune development of the acquired immune system between geladas exposed to takeovers as infants and those who were not. Acquired immune system development is measured by the identification of gastrointestinal (GI) parasites species that commonly infect geladas, as these parasites trigger the adaptive immune response. We identify these GI parasites using high throughput methods to sequence a region of DNA with known variation across nematode species, allowing us to identify parasites at the level of genus. We expect to see greater GI parasite species diversity in geladas with underdeveloped acquired immune systems as we continue to identify GI parasite species in geladas that were exposed to early life adversity and those who were not.

Contextual word representations derived from large-scale neural language models are successful across a diverse set of natural language processing (NLP) tasks, suggesting that they encode useful and transferable features of language. To shed light on the linguistic knowledge they capture, we study the representations produced by several recent pretrained contextualizers (variants of ELMo, the OpenAI transformer LM, and BERT) with a suite of sixteen diverse probing tasks. We find that linear models trained on top of frozen contextual representations are competitive with state-of-the-art task-specific models in many cases, but fail on tasks requiring fine-grained linguistic knowledge (e.g., conjunct identification). To investigate the transferability of contextual word representations, we quantify differences in the transferability of individual layers within contextualizers, especially between recurrent neural networks (RNNs) and transformers. For instance, higher layers of RNNs are more task-specific, while transformer layers do not exhibit the same monotonic trend. In addition, to better understand what makes contextual word representations transferable, we compare language model pretraining with eleven supervised pretraining tasks. For any given task, pretraining on a closely related task yields better performance than language model pretraining (which is better on average) when the pretraining dataset is fixed. However, language model pretraining on more data gives the best results.

Diet influences mammalian physiology and has been linked to obesity and the development of carotid artery disease. Carotid artery disease (CAD) is a result of plaque buildup in arteries leading to the brain and other parts of the head, reducing blood flow and increasing the risk of stroke. Rates of obesity and carotid artery disease are higher in countries with Western diets (higher saturated fat and carbohydrate composition) compared to those with Mediterranean diets (higher produce and monounsaturated fat composition). Although there is evidence demonstrating the effect of diet on overall health and physiology, the mechanisms by which these changes arise could be better understood. These health consequences of diet are thought to be mediated by changes in gene expression caused by eating different diets. Understanding how diet influences gene regulation in different tissue types could give critical insight into our understanding of diet’s effect on physiology. To this end, I measured gene expression via RNA-Seq in 37 cynomolgus macaques (Macaca fascicularis) that were randomly assigned either a Western or Mediterranean diet. By analyzing gene expression in samples collected from two types of fat (subcutaneous and visceral) as well as two arterial tissue types (iliac and carotid), I aim to gain insight into the transcriptomic changes due to diet. Further, by understanding diet’s role in gene regulation, I aim to construct a clearer picture of the mechanisms through which western diet might heighten the risk of obesity and the development of CAD. I predict that the differentially expressed genes in fat tissues will be associated with obesity and that the differentially expressed genes in the arterial tissues will be associated with plaque build up and CAD. This study will illuminate how diet mediates gene regulation in a variety of tissues, providing a mechanistic link to diet-related diseases such as obesity and CAD.