Separate Lines Of Work Research Articles

Loss of HLA-DO function impairs the development of memory CD4 T cells

Abstract DO is an accessory protein that binds DM for trafficking to MIIC and has peptide editing functions. DO is mainly expressed in thymic medulla and B cells. Using biochemical experiments, our lab has discovered that DO has differential effects on editing peptides of different sequences. DO increases binding of DM-resistant peptides, and reduces the binding of DM-sensitive peptides to the HLA-DR1 molecules. In a separate line of work, we have established that reduced densities of antigen presentation by B cells during the contraction phase of an infection, induces quiescence in antigen experienced CD4 T cells as they differentiate into memory T cells. This quiescent phenotype helps memory CD4 T cell survival and promotes effective memory responses to secondary Ag challenge. Based on our mechanistic understanding of DO function, it would be expected that in immunized DO-KO mice, presentation of altered densities of pMHC by B cells would lead to faulty development of memory CD4 T cells. We tracked the development of DR1 restricted H5N1 flu specific memory CD4 T cells and B cells post flu vaccine immunization in DR1+/DO-KO mice and found that immunized DR1+/DO-KO mice had fewer CD4 memory T cells and memory B cells than DR1+/DO-WT mice. By in vitroand in vivorecall experiments, we found that memory responses elicited in DR1+/DO-KO mice were compromised. Using a second model antigen, OVA, we confirmed our findings, i.e., altered memory development and secondary responses in DO-KO mice. These results demonstrate that in the absence of DO, the presentation of cognate foreign antigens in the DO-KO mice is altered and can hamper the proper development of memory cells. These findings suggest a justification for the late appearance of DO in warm blooded mammals. R37 AI060040 R01 AI121174 R01 AI130210

What are the Desired Characteristics of Calibration Sets? Identifying Correlates on Long Form Scientific Summarization.

Summarization models often generate text that is poorly calibrated to quality metrics because they are trained to maximize the likelihood of a single reference (MLE). To address this, recent work has added a calibration step, which exposes a model to its own ranked outputs to improve relevance or, in a separate line of work, contrasts positive and negative sets to improve faithfulness. While effective, much of this work has focused on how to generate and optimize these sets. Less is known about why one setup is more effective than another. In this work, we uncover the underlying characteristics of effective sets. For each training instance, we form a large, diverse pool of candidates and systematically vary the subsets used for calibration fine-tuning. Each selection strategy targets distinct aspects of the sets, such as lexical diversity or the size of the gap between positive and negatives. On three diverse scientific long-form summarization datasets (spanning biomedical, clinical, and chemical domains), we find, among others, that faithfulness calibration is optimal when the negative sets are extractive and more likely to be generated, whereas for relevance calibration, the metric margin between candidates should be maximized and surprise-the disagreement between model and metric defined candidate rankings-minimized. Code to create, select, and optimize calibration sets is available at https://github.com/griff4692/calibrating-summaries.

Academic language and listening comprehension—Two sides of the same coin? An empirical examination of their dimensionality, relations to reading comprehension, and assessment modality.

Two widely studied language skills in relation to reading comprehension are listening comprehension skill and academic language proficiency. Although their constituent skills and theoretical accounts of how they are related to reading comprehension share a large overlap, they have been studied in separate lines of work. In this study, we investigated the dimensionality of listening comprehension and academic language proficiency tasks, their relations to reading comprehension, and the impact of assessment modality (reading vs. oral language) of academic language proficiency, using data from children in Grade 2 (N = 350). Two cohorts of children from the same schools were assessed on the same set of listening comprehension, word reading, reading comprehension, and academic language tasks. Whereas the first three constructs were assessed in identical manner across the 2 cohorts, academic language tasks were assessed in different modalities (1 cohort in a reading context and the other cohort in an oral language context). Academic language proficiency and listening comprehension skill tasks were best described as having a general oral language construct that captured common variance among all the tasks as well as having specific residual factors. Students’ average performance on academic language tasks was lower in the reading context, wherein students’ reading skill was also captured beyond the academic “language” proficiency. Across assessment modalities, it was the general oral language construct, not the specific factors, that was reliable, and consistently and most dominantly related to reading comprehension after accounting for word reading. (PsycInfo Database Record (c) 2020 APA, all rights reserved)

Attentional Bias for Threat and Anxiety: The Role of Loneliness

Objective: There is literature to suggest that anxious individuals may be lonely. Attentional bias for threat (ABT), a mechanism implicated in the core symptoms of anxiety, has been linked to loneliness in a separate line of work. The primary aim of this study was to examine the role of loneliness in the association between ABT and anxiety. Method: An unselected sample of 260 individuals (196 Female; Mean Age = 22.43) completed measures of loneliness, ABT (a dot probe task), and anxiety. Two possible models of the role of loneliness in the ABT-anxiety link were tested using hierarchical regression analysis: (1) A moderation model (the ABT-anxiety link is moderated by loneliness), and (2) A proxy model (the ABT-anxiety link is better explained by loneliness). Results: In support of the latter model, ABT no longer predicted anxiety after the effects of loneliness had been accounted for. Additionally, ABT was associated with anxiety only when indexed using sadness-related scenes (but not fear-related scenes). Conclusions: Loneliness may be one important source of exaggerated threat appraisals which underpin the association between ABT and anxiety. Different classes of negative stimuli may be differentially sensitive to anxiety and should be a point of consideration in future research.

Relative Effects of Maladaptive Traits and Anxiety Mindset on Borderline Personality Disorder Symptoms

Borderline personality disorder (BPD) is a relatively common and debilitating mental health condition. Research has indicated associations between maladaptive personality traits and BPD. A separate line of work suggests that beliefs about how much one's anxiety can change (i.e., anxiety mindset) may influence a wide range of mental health symptoms, many of which people with BPD often exhibit. We begin to integrate these two perspectives in the current study by evaluating the relative effects of maladaptive personality traits and anxiety mindset on BPD symptoms in a large sample of undergraduates (N = 998). Results indicate that traits and mindsets are independent statistical predictors of BPD symptoms. These findings suggest that both maladaptive traits and anxiety mindset are relevant to BPD, which may inform future research on the connections between personality processes and BPD as well as clinical intervention.

Graph-Based Semi-Supervised Learning

While labeled data is expensive to prepare, ever increasing amounts of unlabeled data is becoming widely available. In order to adapt to this phenomenon, several semi-supervised learning (SSL) algorithms, which learn from labeled as well as unlabeled data, have been developed. In a separate line of work, researchers have started to realize that graphs provide a natural way to represent data in a variety of domains. Graph-based SSL algorithms, which bring together these two lines of work, have been shown to outperform the state-of-the-art in many applications in speech processing, computer vision, natural language processing, and other areas of Artificial Intelligence. Recognizing this promising and emerging area of research, this synthesis lecture focuses on graph-based SSL algorithms (e.g., label propagation methods). Our hope is that after reading this book, the reader will walk away with the following: (1) an in-depth knowledge of the current state-of-the-art in graph-based SSL algorithms, and the ability to implement them; (2) the ability to decide on the suitability of graph-based SSL methods for a problem; and (3) familiarity with different applications where graph-based SSL methods have been successfully applied. Table of Contents: Introduction / Graph Construction / Learning and Inference / Scalability / Applications / Future Work / Bibliography / Authors' Biographies / Index

Correcting NDP for SO2 and NOx Emissions: Implementation of a Theoretical Model in Practice

The theoretical and the practical studies in the field of environmental accounting are often two separate lines of work. In this study, we develop an optimal control theory model for adjusting NDP for the effects of SO2 and NOx emissions, and subsequently insert empirically estimated values. The model includes correction entries for the effects on welfare, real capital, health and the quality and quantity of renewable natural resources. In the empirical valuation study, production losses were estimated with dose‐response functions. Recreational and other welfare values were estimated by the contingent valuation (CV) method. Effects on capital depreciation are also included. For comparison, abatement costs and environmental protection expenditures for reducing sulfur and nitrogen emissions were estimated. The theoretical model was then utilized to calculate the adjustment to NDP in a consistent manner. The estimated damage value of sulfur is close to the Swedish sulfur tax.