Search

Statin therapy is a cornerstone in the treatment of systemic vascular diseases. However, statins have failed to translate as therapeutics for pulmonary vascular disease. Early pulmonary vascular disease in the setting of congenital heart disease (CHD) is characterized by endothelial dysfunction, which precedes the more advanced stages of vascular remodeling. These features make CHD an ideal cohort in which to re-evaluate the potential pulmonary vascular benefits of statins, with a focus on endothelial biology. However, it is critical that the full gamut of the pleiotropic effects of statins in the endothelium are uncovered. The purpose of this investigation was to evaluate the therapeutic potential of simvastatin for children with CHD and pulmonary over-circulation, and examine mechanisms of simvastatin action on the endothelium. Our data demonstrate that daily simvastatin treatment preserves endothelial function in our shunt lamb model of pulmonary over-circulation. Further, using pulmonary arterial endothelial cells (PAECs) isolated from Shunt and control lambs, we identified a new mechanism of statin action mediated by increased expression of the endogenous Akt1 inhibitor, C-terminal modifying protein (CTMP). Increases in CTMP were able to decrease the Akt1-mediated mitochondrial redistribution of endothelial nitric oxide synthase (eNOS) which correlated with increased enzymatic coupling, identified by increases in NO generation and decreases in NOS-derived superoxide. Together our data identify a new mechanism by which simvastatin enhances NO signaling in the pulmonary endothelium and identify CTMP as a potential therapeutic target to prevent the endothelial dysfunction that occurs in children born with CHD resulting in pulmonary over-circulation.

INTRODUCTION: This article provides an overview of presentations and discussions from the inaugural Healthcare Delivery Science: Innovation and Partnerships for Health Equity Research (DESCIPHER) Symposium. METHODS: The symposium brought together esteemed experts from various disciplines to explore models for translating evidence-based interventions into practice. RESULTS: The symposium highlighted the importance of disruptive innovation in healthcare, the need for multi-stakeholder engagement, and the significance of family and community involvement in healthcare interventions. CONCLUSIONS: The article concluded with a call to action for advancing healthcare delivery science to achieve health equity.

Diatoms and bacteria play a vital role in investigating the ecological effects of heavy metals in the environment. Despite separate studies on metal interactions with diatoms and bacteria, there is a significant gap in research regarding heavy metal interactions within a diatom-bacterium system, which closely mirrors natural conditions. In this study, we aim to address this gap by examining the interaction of uranium(VI) (U(VI)) with Achnanthidium saprophilum freshwater diatoms and their natural bacterial community, primarily consisting of four successfully isolated bacterial strains (Acidovorax facilis, Agrobacterium fabrum, Brevundimonas mediterranea, and Pseudomonas peli) from the diatom culture. Uranium (U) bio-association experiments were performed both on the xenic A. saprophilum culture and on the four bacterial isolates. Scanning electron microscopy and transmission electron microscopy coupled with spectrum imaging analysis based on energy-dispersive X-ray spectroscopy revealed a clear co-localization of U and phosphorus both on the surface and inside A. saprophilum diatoms and the associated bacterial cells. Time-resolved laser-induced fluorescence spectroscopy with parallel factor analysis identified similar U(VI) binding motifs both on A. saprophilum diatoms and the four bacterial isolates. This is the first work providing valuable microscopic and spectroscopic data on U localization and speciation within a diatom-bacterium system, demonstrating the contribution of the co-occurring bacteria to the overall interaction with U, a factor non-negligible for future modeling and assessment of radiological effects on living microorganisms.

Lanthanides and actinides are emerging contaminants, but little is known about their uptake and distribution by plants and their interactions in the rhizosphere. To better understand the fate of these metals in plants, we assessed the bioassociation of 2, 20 and 200 µM Eu(III) by five hydroponically grown crops endemic to Europe. The metal’s concentration and its speciation was monitored by inductively coupled plasma mass spectrometry and laser spectroscopy, whereas root exudation was investigated by chromatographic methods. It has been shown, that Eu(III) bioassociation is a two-stage process, involving rapid biosorption followed by accumulation in root tissue and distribution to the stem and leaves. Within 96 h of exposure time, the plant induces a change of Eu(III) speciation in the liquid medium, from a predominant Eu(III) aquo species, as calculated by thermodynamic modelling, to a species with longer luminescence lifetime. Root exudates such as citric, malic, and fumaric acid were identified in the cultivation medium and affect Eu(III) speciation in solution, as was shown by a change in the thermodynamic model. These results contribute to a comprehensive understanding of the fate of lanthanides in the biosphere and provide a basis for further investigations with the chemical analogues Cm(III) and Am(III).

Publication of bioassociation, spectroscopic, chromatographic and thermodynamically modelled data obtained in hydroponic plant experiments with Eu(III).

Characterizing the physicochemistry and microbial diversity of U mine water is a key prerequisite for understanding the biogeochemical processes occurring in these water mass and for the design of an efficient bioremediation strategy. This study has collected and analysed in reference date measurements water samples from two former U-mines (Schlema-Alberoda and Pöhla, Wismut GmbH) in East Germany. The samples from both mines are pH-circumneutral (7.3 and 6.6) and show reducing conditions (EH: +139 and –91 mV). Interestingly, the U and sulphate concentrations of Schlema-Alberoda mine water (U: 1 mg/L; SO42−: 335 mg/L) are 2 and 3 order of magnitude higher than those of the Pöhla samples (U: 0.01 mg/L; SO42−: 0.5 mg/L), respectively. U, SO42− and Fe seem to shape the differential microbial diversity of the water from both mines. Microbial diversity analysis revealed the distribution of bacteria with U(VI)-reducing capacity and the ability to maintain the stability of reduced U-species (e.g., Desulfurivibrio, Gallionella and Sulfuricurvum). In addition, water from the mines harbour wood-degrading fungal communities (e.g., composed of Cadophora and Acremonium) providing potential electron donors which promote the growth of U-reducing bacteria. For the design of a bioremediation strategy, we conducted a preliminary U-bioreduction experiment to screen for suitable electron donors (glycerol, vanillic acid and gluconic acid). We also observed that high levels of soluble U (initially present as Ca2UO2(CO3)3(aq) and UO2(CO3)34−), Fe and SO42− were removed by 98, 95 and 53%, respectively from the mine water by using glycerol as electron donor. The remaining U concentrations after bioreduction meet regulatory standards for beneficial reuse of U mine water. As a whole, the results reveal the che

Malware detection is an interesting and valuable domain to work in because it has significant real-world impact and unique machine-learning challenges. We investigate existing long-range techniques and benchmarks and find that they're not very suitable in this problem area. In this paper, we introduce Holographic Global Convolutional Networks (HGConv) that utilize the properties of Holographic Reduced Representations (HRR) to encode and decode features from sequence elements. Unlike other global convolutional methods, our method does not require any intricate kernel computation or crafted kernel design. HGConv kernels are defined as simple parameters learned through backpropagation. The proposed method has achieved new SOTA results on Microsoft Malware Classification Challenge, Drebin, and EMBER malware benchmarks. With log-linear complexity in sequence length, the empirical results demonstrate substantially faster run-time by HGConv compared to other methods achieving far more efficient scaling even with sequence length $\geq 100,000$., Comment: To appear in Proceedings of the 27th International Conference on Artificial Intelligence and Statistics (AISTATS) 2024, Valencia, Spain

As machine learning becomes increasingly prevalent in impactful decisions, recognizing when inference data is outside the model's expected input distribution is paramount for giving context to predictions. Out-of-distribution (OOD) detection methods have been created for this task. Such methods can be split into representation-based or logit-based methods from whether they respectively utilize the model's embeddings or predictions for OOD detection. In contrast to most papers which solely focus on one such group, we address both. We employ dimensionality reduction on feature embeddings in representation-based methods for both time speedups and improved performance. Additionally, we propose DICE-COL, a modification of the popular logit-based method Directed Sparsification (DICE) that resolves an unnoticed flaw. We demonstrate the effectiveness of our methods on the OpenOODv1.5 benchmark framework, where they significantly improve performance and set state-of-the-art results.

While distributed training is often viewed as a solution to optimizing linear models on increasingly large datasets, inter-machine communication costs of popular distributed approaches can dominate as data dimensionality increases. Recent work on non-interactive algorithms shows that approximate solutions for linear models can be obtained efficiently with only a single round of communication among machines. However, this approximation often degenerates as the number of machines increases. In this paper, building on the recent optimal weighted average method, we introduce a new technique, ACOWA, that allows an extra round of communication to achieve noticeably better approximation quality with minor runtime increases. Results show that for sparse distributed logistic regression, ACOWA obtains solutions that are more faithful to the empirical risk minimizer and attain substantially higher accuracy than other distributed algorithms., Comment: Under review

We identify the nonlinear normal modes spawning from the stable equilibrium of a double pendulum under gravity, and we establish their connection to homoclinic orbits through the unstable upright position as energy increases. This result is exploited to devise an efficient swing-up strategy for a double pendulum with weak, saturating actuators. Our approach involves stabilizing the system onto periodic orbits associated with the nonlinear modes while gradually injecting energy. Since these modes are autonomous system evolutions, the required control effort for stabilization is minimal. Even with actuator limitations of less than 1% of the maximum gravitational torque, the proposed method accomplishes the swing-up of the double pendulum by allowing sufficient time., Comment: Preprint of a paper to appear at the European Control Conference (ECC) 2024 in Stockholm, Sweden

Linear models are ubiquitous in data science, but are particularly prone to overfitting and data memorization in high dimensions. To guarantee the privacy of training data, differential privacy can be used. Many papers have proposed optimization techniques for high-dimensional differentially private linear models, but a systematic comparison between these methods does not exist. We close this gap by providing a comprehensive review of optimization methods for private high-dimensional linear models. Empirical tests on all methods demonstrate robust and coordinate-optimized algorithms perform best, which can inform future research. Code for implementing all methods is released online., Comment: 21 pages, 7 figures. To be published at the 2nd IEEE Conference on Secure and Trustworthy Machine Learning (SaTML)

Binary code is pervasive, and binary analysis is a key task in reverse engineering, malware classification, and vulnerability discovery. Unfortunately, while there exist large corpuses of malicious binaries, obtaining high-quality corpuses of benign binaries for modern systems has proven challenging (e.g., due to licensing issues). Consequently, machine learning based pipelines for binary analysis utilize either costly commercial corpuses (e.g., VirusTotal) or open-source binaries (e.g., coreutils) available in limited quantities. To address these issues, we present Assemblage: an extensible cloud-based distributed system that crawls, configures, and builds Windows PE binaries to obtain high-quality binary corpuses suitable for training state-of-the-art models in binary analysis. We have run Assemblage on AWS over the past year, producing 890k Windows PE and 428k Linux ELF binaries across 29 configurations. Assemblage is designed to be both reproducible and extensible, enabling users to publish "recipes" for their datasets, and facilitating the extraction of a wide array of features. We evaluated Assemblage by using its data to train modern learning-based pipelines for compiler provenance and binary function similarity. Our results illustrate the practical need for robust corpuses of high-quality Windows PE binaries in training modern learning-based binary analyses. Assemblage can be downloaded from https://assemblage-dataset.net

Automatic citation generation for sentences in a document or report is paramount for intelligence analysts, cybersecurity, news agencies, and education personnel. In this research, we investigate whether large language models (LLMs) are capable of generating references based on two forms of sentence queries: (a) Direct Queries, LLMs are asked to provide author names of the given research article, and (b) Indirect Queries, LLMs are asked to provide the title of a mentioned article when given a sentence from a different article. To demonstrate where LLM stands in this task, we introduce a large dataset called REASONS comprising abstracts of the 12 most popular domains of scientific research on arXiv. From around 20K research articles, we make the following deductions on public and proprietary LLMs: (a) State-of-the-art, often called anthropomorphic GPT-4 and GPT-3.5, suffers from high pass percentage (PP) to minimize the hallucination rate (HR). When tested with Perplexity.ai (7B), they unexpectedly made more errors; (b) Augmenting relevant metadata lowered the PP and gave the lowest HR; (c) Advance retrieval-augmented generation (RAG) using Mistral demonstrates consistent and robust citation support on indirect queries and matched performance to GPT-3.5 and GPT-4. The HR across all domains and models decreased by an average of 41.93%, and the PP was reduced to 0% in most cases. In terms of generation quality, the average F1 Score and BLEU were 68.09% and 57.51%, respectively; (d) Testing with adversarial samples showed that LLMs, including the Advance RAG Mistral, struggle to understand context, but the extent of this issue was small in Mistral and GPT-4-Preview. Our study contributes valuable insights into the reliability of RAG for automated citation generation tasks., Comment: Work in progress

Herein, Eu(III) was representatively used as luminescent probe to study the chemical environment and to elucidate the molecular interactions of lanthanides with a suspension cell culture of Nicotiana tabacum BY-2. Biochemical methods were combined with luminescence spectroscopy, two-dimensional microspectroscopic mappings, and data deconvolution methods to resolve bioassociation behavior and spatial distribution of Eu(III) in plant cells.

Item does not contain fulltext, BACKGROUND: Nonsyndromic cleft lip with/without cleft palate (nsCL/P) is a congenital malformation of multifactorial etiology. Research has identified >40 genome-wide significant risk loci, which explain less than 40% of nsCL/P heritability. Studies show that some of the hidden heritability is explained by rare penetrant variants. METHODS: To identify new candidate genes, we searched for highly penetrant de novo variants (DNVs) in 50 nsCL/P patient/parent-trios with a low polygenic risk for the phenotype (discovery). We prioritized DNV-carrying candidate genes from the discovery for resequencing in independent cohorts of 1010 nsCL/P patients of diverse ethnicities and 1574 population-matched controls (replication). Segregation analyses and rare variant association in the replication cohort, in combination with additional data (genome-wide association data, expression, protein-protein-interactions), were used for final prioritization. CONCLUSION: In the discovery step, 60 DNVs were identified in 60 genes, including a variant in the established nsCL/P risk gene CDH1. Re-sequencing of 32 prioritized genes led to the identification of 373 rare, likely pathogenic variants. Finally, MDN1 and PAXIP1 were prioritized as top candidates. Our findings demonstrate that DNV detection, including polygenic risk score analysis, is a powerful tool for identifying nsCL/P candidate genes, which can also be applied to other multifactorial congenital malformations.

AIMS: The totality of atherosclerotic plaque derived from coronary computed tomography angiography (CCTA) emerges as a comprehensive measure to assess the intensity of medical treatment that patients need. This study examines the differences in age onset and prognostic significance of atherosclerotic plaque burden between sexes. METHODS AND RESULTS: From a large multi-center CCTA registry the Leiden CCTA score was calculated in 24 950 individuals. A total of 11 678 women (58.5 ± 12.4 years) and 13 272 men (55.6 ± 12.5 years) were followed for 3.7 years for major adverse cardiovascular events (MACE) (death or myocardial infarction). The age where the median risk score was above zero was 12 years higher in women vs. men (64-68 years vs. 52-56 years, respectively, P < 0.001). The Leiden CCTA risk score was independently associated with MACE: score 6-20: HR 2.29 (1.69-3.10); score > 20: HR 6.71 (4.36-10.32) in women, and score 6-20: HR 1.64 (1.29-2.08); score > 20: HR 2.38 (1.73-3.29) in men. The risk was significantly higher for women within the highest score group (adjusted P-interaction = 0.003). In pre-menopausal women, the risk score was equally predictive and comparable with men. In post-menopausal women, the prognostic value was higher for women [score 6-20: HR 2.21 (1.57-3.11); score > 20: HR 6.11 (3.84-9.70) in women; score 6-20: HR 1.57 (1.19-2.09); score > 20: HR 2.25 (1.58-3.22) in men], with a significant interaction for the highest risk group (adjusted P-interaction = 0.004). CONCLUSION: Women developed coronary atherosclerosis approximately 12 years later than men. Post-menopausal women within the highest atherosclerotic burden group were at significantly higher risk for MACE than their male counterparts, which may have implications for the medical treatment intensity.

In a model of congenital heart disease (CHD), we evaluated if chronically increased pulmonary blood flow and pressure were associated with altered respiratory mechanics and gas exchange. Respiratory mechanics and gas exchange were evaluated in 6 shunt, 7 SHAM, and 7 control age-matched lambs. Lambs were anesthetized and mechanically ventilated for 15 min with tidal volume of 10 mL/kg, positive end-expiratory pressure of 5 cmH2O, and inspired oxygen fraction of 0.21. Respiratory system, lung and chest wall compliances (Crs, CL and Ccw, respectively) and resistances (Rrs, RL and Rcw, respectively), and the profile of the elastic pressure-volume curve (%E2) were evaluated. Arterial blood gases and volumetric capnography variables were collected. Comparisons between groups were performed by one-way ANOVA followed by Tukey-Kramer test for normally distributed data and with Kruskal-Wallis test followed by Steel-Dwass test for non-normally distributed data. Average Crs and CL in shunt lambs were 30% and 58% lower than in control, and 56% and 68% lower than in SHAM lambs, respectively. Ccw was 52% and 47% higher and Rcw was 53% and 40% lower in shunt lambs compared to controls and SHAMs, respectively. No difference in %E2 was identified between groups. No difference in respiratory mechanics was observed between control and SHAM lambs. In shunt lambs, Rcw, Crs and CL were decreased and Ccw was increased when compared to control and SHAM lambs. Pulmonary gas exchange did not seem to be impaired in shunt lambs when compared to controls and SHAMs.

In der Umwelt beeinflussen physikalische, chemische und biologische Prozesse das Wanderungsverhalten von langlebigen Radionukliden (RN). Ziel der Forschung der Abteilung Biogeochemie ist es, dominierende Prozesse der Wechselwirkung von Radionukliden in der Biosphäre einschließlich der Nahrungskette zu identifizieren, die Biochemie dieser Prozesse auf molekularer Ebene zu verstehen und ihre Relevanz für Radionuklidmigration und -transfer nicht nur in der Natur, sondern auch im Umfeld eines Endlagers für hochradioaktive Abfälle zuverlässiger abschätzen zu können. Für den sensitiven Nachweis, die Identifikation des Chemismus sowie der Lokalisation von Radionukliden in verschiedenen biologischen Matrices und aquatischen Systemen nutzen wir eine Vielzahl von spektroskopischen und mikroskopischen Verfahren. Eine zentrale Rolle spielt dabei die Laser-induzierte Anregung der Lumineszenz von Actiniden und Lanthaniden, welche in Verbindung mit hochauflösender Mikroskopie die Beschreibung radioökologischer Prozesse in einer neuen Detailtiefe erlaubt. Mit dieser einzigartigen Kombination konnte zum Beispiel eine Actinid-induzierte Stressantwort bei Pflanzenzellen nachgewiesen, und die chemische Bindungsform von Uran in komplexen Umweltproben analysiert werden. Ein wichtiger Teilaspekt ist dabei außerdem die qualitative und quantitative Erfassung der chemo- und radiotoxischen Wirkung endlagerrelevanter als auch natürlich vorkommender Radionuklide (naturally occuring radioactive materials - NORM) in Organismen und ihre Zellen. Weitere Arbeiten befassen sich aktuell mit der Fragestellung, wie sich eine Uranbelastung in der Umgebung von Goldminen auf die im Umland lebende Bevölkerung auswirkt. Dazu wird gegenwärtig in einem gemeinsamen Projekt mit dem VKTA und der Wismut GmbH der Schwermetallgehalt in Haarproben von Personen mittels massenspektrometrischer Methoden bestimmt und bewertet, die im Umland von Johannisburg und nahe des weltweit größten Goldvorkommens, der Witwatersrand

Environmental pollution by heavy metals and radionuclides is one of the biggest challenges which have to be solved globally. In the former U mine of the Wismut GmbH Schlema-Alberoda (Saxony, Germany), mine water is pumped to the surface, where it is treated by a conventional water treatment plant since certain amounts of U (1 mg/L) and other water pollutants can still be found despite remediation by flooding. Microbiological studies of the mine water, using 16S rRNA gene sequencing, revealed a microbial community, which is characterized by a relative abundance of indigenous microbial groups with U(VI)-reduction ability (e.g., sulfur- and iron-oxidizing bacteria and sulfate-reducing bacteria). From previous studies it is known that microbial cycling processes have a significant impact on the complete enzymatic reduction of soluble U(VI) to U(V) and U(IV) by the addition of an electron donor in low U contaminated mine water. In our experiments, a set of anoxic microcosms with mine water were supplemented with glycerol (10mM) as electron donor. The monitoring of the redox potential, pH and the concentration of U, Fe, As and SO42- revealed a substantial decrease of the U(VI) concentration of up to 98%, as well as Fe with up to 91% and SO42- up to 88% after four months, accompanied by a significant change of the redox potential. A thermodynamic Eh-pH dominance diagram calculated using Geochemist's Workbench predicted the reduction of U(VI) and the formation of a solid U(IV)-mineral. The black precipitates, which were formed during the experiments, were analyzed spectroscopically. Extended X-ray absorption fine structure (EXAFS) indicates the formation of immobile uraninite as a U(IV) solid phase. By means of high-energy-resolution fluorescence-detected X-ray absorption near-edge structure (HERFD-XANES) it was even possible to identify U(V), a highly unstable intermediate in the reduction of U(VI) to U(IV) and poorly reported in

The energy transition in Germany is one of the greatest challenges for the coming decades. An important part of this is the phase-out of nuclear power generation and the safe long-term storage of highly radioactive waste in a future deep geological repository. When considering safety aspects of a potential repository, the influence of biological processes on the integrity of a potential repository is increasingly taken into account in addition to the necessary (physical-) chemical data basis. The effects of the bio-factor are manifold. Our research shows that the interaction of radionuclides with whole (micro)organisms or organic substances produced by them form complexes that alter the migration behavior of radionuclides. In addition, the mineral phases of the host rock are altered by bio-degradation and biomineralization processes. Another important aspect is that microbial induced electron transfer processes cause corrosion that reduces the integrity of potential containers. Last but not least, metabolic processes, especially under anaerobic conditions, lead to gas production that directly alters physical parameters such as pressure. However, the knowledge gained will not only be incorporated into state-of-the-art algorithms for modeling hydrogeological and ecological systems, but will also be important far beyond repository research. Studying the interaction of radionuclides with bacteria, for example, has provided new insights that have led to the development of an innovative process for plastic electroplating. The novel process allows the coating of highly complex polymer parts with functional metal surfaces while at the same time substituting highly toxic chemicals and saving resources. The investigation and comparison of different mine waters from former uranium ore mining has also led to the development of a new concept for the cost-effective, biotechnological purification of contaminated mine waters. The process and the potential of the procedure are curre

Deep geological repositories (DGR) will be used for the final disposal of highly radioactive waste. For the safety assessment of the DGR, it is important to consider accident scenarios such as the ingress of water, which might lead to a release of radionuclides (RNs) from the repository into the groundwater. RNs in groundwater can migrate to the surface soil where they can interact with indigenous microorganisms and plants, entering the food chain and posing a health risk to humans. The reliable modelling of the RN uptake into plants requires more than just transfer factors. A more detailed process understanding of the RN uptake into plants, including the effects of soil microorganisms, is necessary. RNs may affect the soil microbial community altering the natural community composition and interactions. Root exudates from plants and microorganisms can alter the speciation of RNs, affecting their bioavailability and mobility. We study the impact of soil microorganisms on the RN transport and uptake into plants at the molecular level. The experiments discussed focus on studying the modulation of soil microbial diversity in the presence of RNs and selected root exudates. In addition, radiation-resistant soil microorganisms will be isolated to study their interaction with RNs and their effect on the degradation of root exudates. This will include characterization of the degradation products and their interactions with RNs. These findings will be used to elaborate radioecological models for the assessment of the RN transport and uptake into the food chain.

The investigation of the interaction of radionuclides with biosystems plays an important role not only in repository research but also in radioecology. The obtained knowledge expands the basic understanding of the behavior of radionuclides in nature on the one hand and opens up possibilities for the industrial use of the processes on the other hand. The lecture will give a general overview of the research work in biogeochemistry and its application potential.

The aim of the lecture is to give an overview of the equipment at the Institute of Resource Ecology and what analytical options this includes. Furthermore, current research work is presented in the context of the remediation of contaminated water and soil. Building on this, the potential and limits of biological processes and possibilities for collaboration will be shown.

Herein, Eu(III) was representatively used as luminescent probe to study the chemical environment and to elucidate the molecular interactions of lanthanides with a suspension cell culture of Nicotiana tabacum BY-2. Biochemical methods were combined with luminescence spectroscopy, two-dimensional microspectroscopic mappings, and data deconvolution methods to resolve bioassociation behavior and spatial distribution of Eu(III) in plant cells.

Pseudomonas sp. are indigenous inhabitants of ombrotrophic bogs which can survive in acidic, nutrient-poor environments with wide temperature fluctuations. Their interactions with contaminant radionuclides can influence radionuclide biogeochemistry in boreal environment. Here, uranium (U(VI)) bioassociation by Pseudomonas sp. PS-0-L, V4-5-SB and T5-6-I isolated from a boreal bog was studied by a combination of batch contact experiments, spectroscopy and microscopy. All strains removed U from the solution and the U bioassociation efficiency was affected by the nutrient source, incubation temperature, time and pH. Highest U bioassociation occurred in the strains PS-0-L (0.199 mg U/gBDW) and V4-5-SB (0.223 mg U/gBDW). Based on in-situ attenuated total reflection Fourier transformation infrared spectroscopy (ATR FT-IR) analyses, the most likely functional groups responsible for U binding were the cell surface carboxyl groups. In addition, transmission electron microscopy with energy dispersive X-ray spectroscopy (TEM/EDX) showed dense intra-cellular round- and needle-like U accumulations in the cytoplasm and near to the inner cell membrane. The presence of U with phosphorus was indicated in elemental mapping. Modelled data showed ≡SOOHx-1 and ≡SOCO2Hx-1 as the dominant surface sites, contributing to the negative cell surface charge. The U removal efficiency depended on the U(VI) speciation under different pH conditions. At pH 5, the main species reacting with bacterial cell surfaces was UO22+, while at pH 9 UO2(OH)2 and UO2(OH)3- dominated the reactions. Further, U bioassociation increased with increasing aqueous U(VI) concentrations. Our data suggests U bioassociation on 1) outer cell membrane/cell wall associated carboxyl groups (e.g., proteins), and 2) intracellular phosphate groups (e.g., phospholipids).

Environmental pollution by heavy metals and radionuclides is one of the biggest challenges which have to be solved globally. In the former U mine of the Wismut GmbH Schlema-Alberoda (Saxony, Germany), mine water is pumped to the surface, where it is treated by a conventional water treatment plant since certain amounts of U (1 mg/L) and other water pollutants can still be found despite remediation by flooding. Microbiological studies of the mine water, using 16S rRNA gene sequencing, revealed a microbial community, which is characterized by a relative abundance of indigenous microbial groups with U(VI)-reduction ability (e.g., sulfur- and iron-oxidizing bacteria and sulfate-reducing bacteria). From previous studies it is known that microbial cycling processes have a significant impact on the complete enzymatic reduction of soluble U(VI) to U(V) and U(IV) by the addition of an electron donor in low U contaminated mine water. In our experiments, a set of anoxic microcosms with mine water were supplemented with glycerol (10mM) as electron donor. The monitoring of the redox potential, pH and the concentration of U, Fe, As and SO42- revealed a substantial decrease of the U(VI) concentration of up to 98%, as well as Fe with up to 91% and SO42- up to 88% after four months, accompanied by a significant change of the redox potential. A thermodynamic Eh-pH dominance diagram calculated using Geochemist's Workbench predicted the reduction of U(VI) and the formation of a solid U(IV)-mineral. The black precipitates, which were formed during the experiments, were analyzed spectroscopically. Extended X-ray absorption fine structure (EXAFS) indicates the formation of immobile uraninite as a U(IV) solid phase. By means of high-energy-resolution fluorescence-detected X-ray absorption near-edge structure (HERFD-XANES) it was even possible to identify U(V), a highly unstable intermediate in the reduction of U(VI) to U(IV) and poorly reported in

Deep geological repositories (DGR) will be used for the final disposal of highly radioactive waste. For the safety assessment of the DGR, it is important to consider accident scenarios such as the ingress of water, which might lead to a release of radionuclides (RNs) from the repository into the groundwater. RNs in groundwater can migrate to the surface soil where they can interact with indigenous microorganisms and plants, entering the food chain and posing a health risk to humans. The reliable modelling of the RN uptake into plants requires more than just transfer factors. A more detailed process understanding of the RN uptake into plants, including the effects of soil microorganisms, is necessary. RNs may affect the soil microbial community altering the natural community composition and interactions. Root exudates from plants and microorganisms can alter the speciation of RNs, affecting their bioavailability and mobility. We study the impact of soil microorganisms on the RN transport and uptake into plants at the molecular level. The experiments discussed focus on studying the modulation of soil microbial diversity in the presence of RNs and selected root exudates. In addition, radiation-resistant soil microorganisms will be isolated to study their interaction with RNs and their effect on the degradation of root exudates. This will include characterization of the degradation products and their interactions with RNs. These findings will be used to elaborate radioecological models for the assessment of the RN transport and uptake into the food chain.

The energy transition in Germany is one of the greatest challenges for the coming decades. An important part of this is the phase-out of nuclear power generation and the safe long-term storage of highly radioactive waste in a future deep geological repository. When considering safety aspects of a potential repository, the influence of biological processes on the integrity of a potential repository is increasingly taken into account in addition to the necessary (physical-) chemical data basis. The effects of the bio-factor are manifold. Our research shows that the interaction of radionuclides with whole (micro)organisms or organic substances produced by them form complexes that alter the migration behavior of radionuclides. In addition, the mineral phases of the host rock are altered by bio-degradation and biomineralization processes. Another important aspect is that microbial induced electron transfer processes cause corrosion that reduces the integrity of potential containers. Last but not least, metabolic processes, especially under anaerobic conditions, lead to gas production that directly alters physical parameters such as pressure. However, the knowledge gained will not only be incorporated into state-of-the-art algorithms for modeling hydrogeological and ecological systems, but will also be important far beyond repository research. Studying the interaction of radionuclides with bacteria, for example, has provided new insights that have led to the development of an innovative process for plastic electroplating. The novel process allows the coating of highly complex polymer parts with functional metal surfaces while at the same time substituting highly toxic chemicals and saving resources. The investigation and comparison of different mine waters from former uranium ore mining has also led to the development of a new concept for the cost-effective, biotechnological purification of contaminated mine waters. The process and the potential of the procedure are curre

In der Umwelt beeinflussen physikalische, chemische und biologische Prozesse das Wanderungsverhalten von langlebigen Radionukliden (RN). Ziel der Forschung der Abteilung Biogeochemie ist es, dominierende Prozesse der Wechselwirkung von Radionukliden in der Biosphäre einschließlich der Nahrungskette zu identifizieren, die Biochemie dieser Prozesse auf molekularer Ebene zu verstehen und ihre Relevanz für Radionuklidmigration und -transfer nicht nur in der Natur, sondern auch im Umfeld eines Endlagers für hochradioaktive Abfälle zuverlässiger abschätzen zu können. Für den sensitiven Nachweis, die Identifikation des Chemismus sowie der Lokalisation von Radionukliden in verschiedenen biologischen Matrices und aquatischen Systemen nutzen wir eine Vielzahl von spektroskopischen und mikroskopischen Verfahren. Eine zentrale Rolle spielt dabei die Laser-induzierte Anregung der Lumineszenz von Actiniden und Lanthaniden, welche in Verbindung mit hochauflösender Mikroskopie die Beschreibung radioökologischer Prozesse in einer neuen Detailtiefe erlaubt. Mit dieser einzigartigen Kombination konnte zum Beispiel eine Actinid-induzierte Stressantwort bei Pflanzenzellen nachgewiesen, und die chemische Bindungsform von Uran in komplexen Umweltproben analysiert werden. Ein wichtiger Teilaspekt ist dabei außerdem die qualitative und quantitative Erfassung der chemo- und radiotoxischen Wirkung endlagerrelevanter als auch natürlich vorkommender Radionuklide (naturally occuring radioactive materials - NORM) in Organismen und ihre Zellen. Weitere Arbeiten befassen sich aktuell mit der Fragestellung, wie sich eine Uranbelastung in der Umgebung von Goldminen auf die im Umland lebende Bevölkerung auswirkt. Dazu wird gegenwärtig in einem gemeinsamen Projekt mit dem VKTA und der Wismut GmbH der Schwermetallgehalt in Haarproben von Personen mittels massenspektrometrischer Methoden bestimmt und bewertet, die im Umland von Johannisburg und nahe des weltweit größten Goldvorkommens, der Witwatersrand

The presentation gives an overview of the different microbial diets and how plants, fungi and bacteria can be used for the bioremediation of uranium contaminated sites and waters. In particular, the advantages and disadvantages of the various approaches are named and a recommendation is given.

The BLOOM model is a large publicly available multilingual language model, but its pretraining was limited to 46 languages. To extend the benefits of BLOOM to other languages without incurring prohibitively large costs, it is desirable to adapt BLOOM to new languages not seen during pretraining. In this work, we apply existing language adaptation strategies to BLOOM and benchmark its zero-shot prompting performance on eight new languages in a resource-constrained setting. We find language adaptation to be effective at improving zero-shot performance in new languages. Surprisingly, we find that adapter-based finetuning is more effective than continued pretraining for large models. In addition, we discover that prompting performance is not significantly affected by language specifics, such as the writing system. It is primarily determined by the size of the language adaptation data. We also add new languages to BLOOMZ, which is a multitask finetuned version of BLOOM capable of following task instructions zero-shot. We find including a new language in the multitask fine-tuning mixture to be the most effective method to teach BLOOMZ a new language. We conclude that with sufficient training data language adaptation can generalize well to diverse languages. Our code is available at https://github.com/bigscience-workshop/multilingual-modeling. © 2023 Association for Computational Linguistics.

[Background]: Nonsyndromic cleft lip with/without cleft palate (nsCL/P) is a congenital malformation of multifactorial etiology. Research has identified >40 genome-wide significant risk loci, which explain less than 40% of nsCL/P heritability. Studies show that some of the hidden heritability is explained by rare penetrant variants. [Methods]: To identify new candidate genes, we searched for highly penetrant de novo variants (DNVs) in 50 nsCL/P patient/parent-trios with a low polygenic risk for the phenotype (discovery). We prioritized DNV-carrying candidate genes from the discovery for resequencing in independent cohorts of 1010 nsCL/P patients of diverse ethnicities and 1574 population-matched controls (replication). Segregation analyses and rare variant association in the replication cohort, in combination with additional data (genome-wide association data, expression, protein–protein-interactions), were used for final prioritization. [Conclusion]: In the discovery step, 60 DNVs were identified in 60 genes, including a variant in the established nsCL/P risk gene CDH1. Re-sequencing of 32 prioritized genes led to the identification of 373 rare, likely pathogenic variants. Finally, MDN1 and PAXIP1 were prioritized as top candidates. Our findings demonstrate that DNV detection, including polygenic risk score analysis, is a powerful tool for identifying nsCL/P candidate genes, which can also be applied to other multifactorial congenital malformations.

The principal nature-based solution for offsetting relative sea-level rise in the Ganges-Brahmaputra delta is the unabated delivery, dispersal, and deposition of the rivers’ ~1 billion-tonne annual sediment load. Recent hydrological transport modeling suggests that strengthening monsoon precipitation in the 21st century could increase this sediment delivery 34-60%; yet other studies demonstrate that sediment could decline 15-80% if planned dams and river diversions are fully implemented. We validate these modeled ranges by developing a comprehensive field-based sediment budget that quantifies the supply of Ganges-Brahmaputra river sediment under varying Holocene climate conditions. Our data reveal natural responses in sediment supply comparable to previously modeled results and suggest that increased sediment delivery may be capable of offsetting accelerated sea-level rise. This prospect for a naturally sustained Ganges-Brahmaputra delta presents possibilities beyond the dystopian future often posed for this system, but the implementation of currently proposed dams and diversions would preclude such opportunities.

BackgroundNonsyndromic cleft lip with/without cleft palate (nsCL/P) is a congenital malformation of multifactorial etiology. Research has identified >40 genome-wide significant risk loci, which explain less than 40% of nsCL/P heritability. Studies show that some of the hidden heritability is explained by rare penetrant variants. MethodsTo identify new candidate genes, we searched for highly penetrant de novo variants (DNVs) in 50 nsCL/P patient/parent-trios with a low polygenic risk for the phenotype (discovery).We prioritized DNV-carrying candidate genes from the discovery for resequencing in independent cohorts of 1010 nsCL/P patients of diverse ethnicities and 1574 population-matched controls (replication). Segregation analyses and rare variant association in the replication cohort, in combination with additional data (genome-wide association data, expression, protein-protein-interactions), were used for final prioritization. ConclusionIn the discovery step, 60 DNVs were identified in 60 genes, including a variant in the established nsCL/P risk gene CDH1. Re-sequencing of 32 prioritized genes led to the identification of 373 rare, likely pathogenic variants. Finally, MDN1 and PAXIP1 were prioritized as top candidates. Our findings demonstrate that DNV detection, including polygenic risk score analysis, is a powerful tool for identifying nsCL/P candidate genes, which can also be applied to other multifactorial congenital malformations.

Over the past decade, the machine learning security community has developed a myriad of defenses for evasion attacks. An understudied question in that community is: for whom do these defenses defend? This work considers common approaches to defending learned systems and how security defenses result in performance inequities across different sub-populations. We outline appropriate parity metrics for analysis and begin to answer this question through empirical results of the fairness implications of machine learning security methods. We find that many methods that have been proposed can cause direct harm, like false rejection and unequal benefits from robustness training. The framework we propose for measuring defense equality can be applied to robustly trained models, preprocessing-based defenses, and rejection methods. We identify a set of datasets with a user-centered application and a reasonable computational cost suitable for case studies in measuring the equality of defenses. In our case study of speech command recognition, we show how such adversarial training and augmentation have non-equal but complex protections for social subgroups across gender, accent, and age in relation to user coverage. We present a comparison of equality between two rejection-based defenses: randomized smoothing and neural rejection, finding randomized smoothing more equitable due to the sampling mechanism for minority groups. This represents the first work examining the disparity in the adversarial robustness in the speech domain and the fairness evaluation of rejection-based defenses., Comment: Accepted to AISec'23

Thermal facial imagery offers valuable insight into physiological states such as inflammation and stress by detecting emitted radiation in the infrared spectrum, which is unseen in the visible spectra. Telemedicine applications could benefit from thermal imagery, but conventional computers are reliant on RGB cameras and lack thermal sensors. As a result, we propose the Visible-to-Thermal Facial GAN (VTF-GAN) that is specifically designed to generate high-resolution thermal faces by learning both the spatial and frequency domains of facial regions, across spectra. We compare VTF-GAN against several popular GAN baselines and the first conditional Denoising Diffusion Probabilistic Model (DDPM) for VT face translation (VTF-Diff). Results show that VTF-GAN achieves high quality, crisp, and perceptually realistic thermal faces using a combined set of patch, temperature, perceptual, and Fourier Transform losses, compared to all baselines including diffusion.

Subsampling algorithms are a natural approach to reduce data size before fitting models on massive datasets. In recent years, several works have proposed methods for subsampling rows from a data matrix while maintaining relevant information for classification. While these works are supported by theory and limited experiments, to date there has not been a comprehensive evaluation of these methods. In our work, we directly compare multiple methods for logistic regression drawn from the coreset and optimal subsampling literature and discover inconsistencies in their effectiveness. In many cases, methods do not outperform simple uniform subsampling., Comment: To appear in the Thirty-Seventh AAAI Conference on Artificial Intelligence

Concept erasure aims to remove specified features from a representation. It can improve fairness (e.g. preventing a classifier from using gender or race) and interpretability (e.g. removing a concept to observe changes in model behavior). We introduce LEAst-squares Concept Erasure (LEACE), a closed-form method which provably prevents all linear classifiers from detecting a concept while changing the representation as little as possible, as measured by a broad class of norms. We apply LEACE to large language models with a novel procedure called "concept scrubbing," which erases target concept information from every layer in the network. We demonstrate our method on two tasks: measuring the reliance of language models on part-of-speech information, and reducing gender bias in BERT embeddings. Code is available at https://github.com/EleutherAI/concept-erasure.

In recent years, self-attention has become the dominant paradigm for sequence modeling in a variety of domains. However, in domains with very long sequence lengths the $\mathcal{O}(T^2)$ memory and $\mathcal{O}(T^2 H)$ compute costs can make using transformers infeasible. Motivated by problems in malware detection, where sequence lengths of $T \geq 100,000$ are a roadblock to deep learning, we re-cast self-attention using the neuro-symbolic approach of Holographic Reduced Representations (HRR). In doing so we perform the same high-level strategy of the standard self-attention: a set of queries matching against a set of keys, and returning a weighted response of the values for each key. Implemented as a ``Hrrformer'' we obtain several benefits including $\mathcal{O}(T H \log H)$ time complexity, $\mathcal{O}(T H)$ space complexity, and convergence in $10\times$ fewer epochs. Nevertheless, the Hrrformer achieves near state-of-the-art accuracy on LRA benchmarks and we are able to learn with just a single layer. Combined, these benefits make our Hrrformer the first viable Transformer for such long malware classification sequences and up to $280\times$ faster to train on the Long Range Arena benchmark. Code is available at \url{https://github.com/NeuromorphicComputationResearchProgram/Hrrformer}, Comment: To appear in Proceedings of the 40th International Conference on Machine Learning (ICML)

LASSO regularized logistic regression is particularly useful for its built-in feature selection, allowing coefficients to be removed from deployment and producing sparse solutions. Differentially private versions of LASSO logistic regression have been developed, but generally produce dense solutions, reducing the intrinsic utility of the LASSO penalty. In this paper, we present a differentially private method for sparse logistic regression that maintains hard zeros. Our key insight is to first train a non-private LASSO logistic regression model to determine an appropriate privatized number of non-zero coefficients to use in final model selection. To demonstrate our method's performance, we run experiments on synthetic and real-world datasets., Comment: 20 pages, 5 figures

Memorization, or the tendency of large language models (LLMs) to output entire sequences from their training data verbatim, is a key concern for safely deploying language models. In particular, it is vital to minimize a model's memorization of sensitive datapoints such as those containing personal identifiable information (PII). The prevalence of such undesirable memorization can pose issues for model trainers, and may even require discarding an otherwise functional model. We therefore seek to predict which sequences will be memorized before a large model's full train-time by extrapolating the memorization behavior of lower-compute trial runs. We measure memorization of the Pythia model suite and plot scaling laws for forecasting memorization, allowing us to provide equi-compute recommendations to maximize the reliability (recall) of such predictions. We additionally provide further novel discoveries on the distribution of memorization scores across models and data. We release all code and data necessary to reproduce the results in this paper at https://github.com/EleutherAI/pythia

How do large language models (LLMs) develop and evolve over the course of training? How do these patterns change as models scale? To answer these questions, we introduce \textit{Pythia}, a suite of 16 LLMs all trained on public data seen in the exact same order and ranging in size from 70M to 12B parameters. We provide public access to 154 checkpoints for each one of the 16 models, alongside tools to download and reconstruct their exact training dataloaders for further study. We intend \textit{Pythia} to facilitate research in many areas, and we present several case studies including novel results in memorization, term frequency effects on few-shot performance, and reducing gender bias. We demonstrate that this highly controlled setup can be used to yield novel insights toward LLMs and their training dynamics. Trained models, analysis code, training code, and training data can be found at \url{https://github.com/EleutherAI/pythia}., Comment: Code at https://github.com/EleutherAI/pythia

For a class of biped robots with impulsive dynamics and a non-empty set of passive gaits (unactuated, periodic motions of the biped model), we present a method for computing continuous families of locally optimal gaits with respect to a class of commonly used energetic cost functions (e.g., the integral of torque-squared). We compute these families using only the passive gaits of the biped, which are globally optimal gaits with respect to these cost functions. Our approach fills in an important gap in the literature when computing a library of locally optimal gaits, which often do not make use of these globally optimal solutions as seed values. We demonstrate our approach on a well-studied two-link biped model., Comment: Published in the IROS 2023 proceedings. This version has several updated sections and figures. The proceedings contains a supplementary video, see https://ieeexplore.ieee.org/document/10342322/media#media . An implementation of the results are also available, see https://github.com/nr-codes/OptimalGaitsForCompassGait