Search

Machine learning algorithms and neural networks have recently been used for the classification of middle ear disorders using wideband acoustic immittance and wideband tympanometry data. This study applies the extreme gradient boosting (XGB) classifier, trained on simulated WAI data, to classify real measured data for normal, otosclerotic, and disarticulated ears. The achieved macro recall of 82 % is comparable to other approaches trained with real measurement data. The interpretability methods LIME and SHAP are used to quantify each feature’s contribution, both revealing energy reflectance between 600-800 Hz as a key feature for all classes. The key feature identified matches the differences that can be visually observed in the training and test data. However, the obtained feature contributions don’t provide enough distinguishable information to recognise incorrect or uncertain classifications.

Objectives: To demonstrate how librarians can use engagement strategies to foster the exchange of knowledge and skills for data analysis and to build bridges between data communities. A second objective is to help student instructors to develop effective live-coding pedagogical practices and to gain practical experience in leading participatory workshop sessions. Methods: Librarians developed a low-barrier introductory peer-to-peer data science workshop series to support students seeking to develop coding, data analysis, and visualization skills, with a focus on Python and SQL. We guided undergraduate peer instructors in participatory live-coding pedagogy, organized practice sessions for instructors, and managed the scheduling, logistics, outreach, and hosting of the workshops. Results: In Fall 2023 sessions in the workshop series were delivered synchronously to over 100 participants, including students from our home institution and more than a dozen community colleges; one workshop was delivered twice—once in English, once in Spanish. Workshop recordings posted online have been viewed over 1000 times. Conclusions: We successfully identified strategies for building upon existing relationships and strengthening connections among diverse data communities; designing programs and outreach efforts to lower barriers to participation in data science; and fostering a culture of diversity, equity, and inclusion in data science knowledge sharing.

The application of force-velocity (FV) profiling in American football has yet to be explored. Purpose: To measure and compare FV profiles in collegiate American football players grouped by position, and to determine if FV profiles could predict countermovement jump (CMJ) height and sprinting performance. Methods: Horizontal and vertical FV profiles, CMJ and sprinting performance were assessed in 81 collegiate American football players. One-way ANOVAs were used to determine if significant differences in FV profiles existed between position groups (big: offensive/defensive lineman, big skill: linebacker, skill: quarterback). Correlation analyses were used to determine if performance measures (CMJ, sprints) were related to FV profiles (maximum force, F0, maximum velocity, V0, maximum power, Pmax). We hypothesized that 1) “big” athletes would have the highest F0, and 2) horizontal and vertical FV profiling metrics would correlate with sprinting performance and CMJ height, respectively. Results: “Big” athletes had the highest absolute F0 in the horizontal FV profiles but when normalized to body weight, they had the lowest F0 and 77% were classified as force-deficient. When accounting for body weight, vertical FV metrics explained 62.8% of the variance in CMJ height and horizontal FV metrics accounted for 85.0% of the variance in sprinting performance. Conclusion: Athletes’ FV imbalance could not be predicted by their position. Vertical- and horizontal-related FV variables predicted performance metrics that were performed in the same plane, suggesting that FV profiling could be a useful performance assessment tool in American football.

Inflammation plays a pivotal role in the pathogenesis of primary and post-essential thrombocythemia or post-polycythemia vera myelofibrosis (MF) in close cooperation with the underlying molecular drivers. This inflammatory state is induced by a dynamic spectrum of inflammatory cytokines, although recent evidence points to the participation of additional soluble inflammatory mediators. Damage-associated molecular patterns (DAMPs) represent endogenous signals released upon cell death or damage which trigger a potent innate immune response. We assessed the contribution of two prototypical DAMPs, HMGB1 and S100A8/A9, to MF inflammation. Circulating HMGB1 and S100A8/A9 were elevated in MF patients in parallel to the degree of systemic inflammation and levels increased progressively during advanced disease stages. Patients with elevated DAMPs had higher frequency of adverse clinical features, such as anemia, and inferior survival, suggesting their contribution to disease progression. Monocytes, which are key players in MF inflammation, were identified as a source of S100A8/A9 but not HMGB1 release, while both DAMPs correlated with cell death parameters, such as serum LDH and cell-free DNA, indicating that passive release is an additional mechanism leading to increased DAMPs. HMGB1 and S100A8/A9 promote inflammation through binding to Toll-like receptor (TLR) 4, whereas the former also binds TLR2. Monocytes from MF patients were shown to be hyperactivated at baseline, as reflected by higher CD11b and tissue factor exposure and increased expression levels of proinflammatory cytokines IL-1β and IL-6. Patient monocytes showed preserved TLR4 and TLR2 expression and were able to mount normal or even exacerbated functional responses and cytokine upregulation following stimulation of TLR4 and TLR2. Elevated levels of endogenous TLR ligands HMGB1 and S100A8/A9 coupled to the finding of preserved or hyperreactive TLR-triggered responses indicate that DAMPs may promote monocyte activation and cytokine production in MF, fueling inflammation. Plasma IL-1β and IL-6 were elevated in MF and correlated with DAMPs levels, raising the possibility that DAMPs could contribute to cytokine generation in vivo. In conclusion, this study highlights that, in cooperation with classic proinflammatory cytokines, DAMPs represent additional inflammatory mediators that may participate in the generation of MF inflammatory state, potentially providing novel biomarkers of disease progression and new therapeutic targets.

DNA computing is a highly interdisciplinary field which combines molecular operations with theoretical algorithm design. A number of algorithms have been demonstrated in DNA computing, but to date network flow problems have not been studied. We aim to provide an approach to calculate the value of the maximum flow in networks by encoding the mathematical problem in DNA molecules and by using molecular biology techniques to manipulate the DNA. We present results which demonstrate that the algorithm works for an example network problem.

Student employees at circulation desks are the first points of contact for many library users with research questions, but they are challenged to respond or refer when librarians are unavailable. In an attempt to cultivate student employees' reference skills, confidence, and connection to the library's values and mission, the Engineering and Physical Sciences Division of the UC Berkeley Library developed an outcomes-based training program consisting of an interactive session and a dynamic online reference manual. Student employees completed pre-training and post-training assessments of their reference skills and self-efficacy. Qualitative feedback was analyzed to identify ongoing challenges and intangible benefits for student employees engaged in reference work. Challenges observed include user expectations, infrequent questions, and the complex information landscape; benefits observed include community building and student growth. Implications of these challenges and benefits for student training programs are discussed.

Myelofibrosis (MF) is a clonal hematopoietic stem cell disorder classified among chronic myeloproliferative neoplasms, characterized by exacerbated myeloid and megakaryocytic proliferation and bone marrow fibrosis. It is induced by driver (JAK2/CALR/MPL) and high molecular risk mutations coupled to a sustained inflammatory state that contributes to disease pathogenesis. Patient outcome is determined by stratification into risk groups and refinement of current prognostic systems may help individualize treatment decisions. Circulating cell-free (cf)DNA comprises short fragments of double-stranded DNA, which promotes inflammation by stimulating several pathways, including inflammasome activation, which is responsible for IL-1β and IL-18 maturation and release. In this work, we assessed the contribution of cfDNA as a marker of disease progression and mediator of inflammation in MF. cfDNA was increased in MF patients and higher levels were associated with adverse clinical outcome, a high-risk molecular profile, advanced disease stages and inferior overall survival, indicating its potential value as a prognostic marker. Cell-free DNA levels correlated with tumor burden parameters and markers of systemic inflammation. To mimic the effects of cfDNA, monocytes were stimulated with poly(dA:dT), a synthetic double-stranded DNA. Following stimulation, patient monocytes released higher amounts of inflammasome-processed cytokine, IL-18 to the culture supernatant, reflecting enhanced inflammasome function. Despite overexpression of cytosolic DNA inflammasome sensor AIM2, IL-18 release from MF monocytes was shown to rely mainly on the NLRP3 inflammasome, as it was prevented by NLRP3-specific inhibitor MCC950. Circulating IL-18 levels were increased in MF plasma, reflecting in vivo inflammasome activation, and highlighting the previously unrecognized involvement of this cytokine in MF cytokine network. Monocyte counts were higher in patients and showed a trend towards correlation with IL-18 levels, suggesting monocytes represent a source of circulating IL-18. The close correlation shown between IL-18 and cfDNA levels, together with the finding of enhanced DNA-triggered IL-18 release from monocytes, suggest that cfDNA promotes inflammation, at least in part, through inflammasome activation. This work highlights cfDNA, the inflammasome and IL-18 as additional players in the complex inflammatory circuit that fosters MF progression, potentially providing new therapeutic targets.

Beim Erlass des Arbeitsgesetzes vor bald sechzig Jahren wurden Arbeitnehmende, die eine wissenschaftliche Tätigkeit ausüben, vom Geltungsbereich ausgenommen. Eine wissenschaftliche Karriere ist heute oft mit prekären Arbeitsbedingungen verbunden. Daher stellt sich die Frage, ob es gerechtfertigt war und ist, diese Arbeitnehmenden generell vom Schutzbereich auszunehmen. Der vorliegende Beitrag geht dieser Frage anhand der Entstehungsgeschichte nach und analysiert die Situation aus der heutigen Perspektive. -- Il y a près de soixante ans, le législateur décidait d'exclure les travailleurs scientifiques du champ d'application de la loi fédérale sur le travail. Aujourd'hui, les carrières scientifiques sont souvent liées à des conditions de travail précaires. La question se pose de savoir s'il était, et s'il est toujours, justifié d'exclure de manière générale ce type de travailleurs du champ de protection de la loi. Pour y répondre, l'article examine l'historicité de la loi sur le travail et analyse la situation d'un point de vue actuel.