Modelling dataset bias in machine-learned theories of economic decision-making

Normative and descriptive models have long vied to explain and predict human risky choices, such as those between goods or gambles. A recent study reported the discovery of a new, more accurate model of human decision-making by training neural networks on a new online large-scale dataset, choices13k. Here we systematically analyse the relationships between several models and datasets using machine-learning methods and find evidence for dataset bias. Because participants' choices in stochastically dominated gambles were consistently skewed towards equipreference in the choices13k dataset, we hypothesized that this reflected increased decision noise. Indeed, a probabilistic generative model adding structured decision noise to a neural network trained on data from a laboratory study transferred best, that is, outperformed all models apart from those trained on choices13k. We conclude that a careful combination of theory and data analysis is still required to understand the complex interactions of machine-learning models and data of human risky choices.

Role of compliant mechanics and motor control in hopping - from human to robot

Compliant leg function found during bouncy gaits in humans and animals can be considered a role model for designing and controlling bioinspired robots and assistive devices. The human musculoskeletal design and control differ from distal to proximal joints in the leg. The specific mechanical properties of different leg parts could simplify motor control, e.g., by taking advantage of passive body dynamics. This control embodiment is complemented by neural reflex circuitries shaping human motor control. This study investigates the contribution of specific passive and active properties at different leg joint levels in human hopping at different hopping frequencies. We analyze the kinematics and kinetics of human leg joints to design and control a bioinspired hopping robot. In addition, this robot is used as a test rig to validate the identified concepts from human hopping. We found that the more distal the joint, the higher the possibility of benefit from passive compliant leg structures. A passive elastic element nicely describes the ankle joint function. In contrast, a more significant contribution to energy management using an active element (e.g., by feedback control) is predicted for the knee and hip joints. The ankle and knee joints are the key contributors to adjusting hopping frequency. Humans can speed up hopping by increasing ankle stiffness and tuning corresponding knee control parameters. We found that the force-modulated compliance (FMC) as an abstract reflex-based control beside a fixed spring can predict human knee torque-angle patterns at different frequencies. These developed bioinspired models for ankle and knee joints were applied to design and control the EPA-hopper-II robot. The experimental results support our biomechanical findings while indicating potential robot improvements. Based on the proposed model and the robot's experimental results, passive compliant elements (e.g. tendons) have a larger capacity to contribute to the distal joint function compared to proximal joints. With the use of more compliant elements in the distal joint, a larger contribution to managing energy changes is observed in the upper joints.

A reduced SNP panel optimised for non-invasive genetic assessment of a genetically impoverished conservation icon, the European bison

The European bison was saved from the brink of extinction due to considerable conservation efforts since the early twentieth century. The current global population of > 9500 individuals is the result of successful ex situ breeding based on a stock of only 12 founders, resulting in an extremely low level of genetic variability. Due to the low allelic diversity, traditional molecular tools, such as microsatellites, fail to provide sufficient resolution for accurate genetic assessments in European bison, let alone from non-invasive samples. Here, we present a SNP panel for accurate high-resolution genotyping of European bison, which is suitable for a wide variety of sample types. The panel accommodates 96 markers allowing for individual and parental assignment, sex determination, breeding line discrimination, and cross-species detection. Two applications were shown to be utilisable in further Bos species with potential conservation significance. The new SNP panel will allow to tackle crucial tasks in European bison conservation, including the genetic monitoring of reintroduced populations, and a molecular assessment of pedigree data documented in the world's first studbook of a threatened species.

Cytokine-responsive T- and NK-cells portray SARS-CoV-2 vaccine-responders and infection in multiple myeloma patients

Patients with multiple myeloma (MM) routinely receive mRNA-based vaccines to reduce COVID-19-related mortality. However, whether disease- and therapy-related alterations in immune cells and cytokine-responsiveness contribute to the observed heterogeneous vaccination responses is unclear. Thus, we analyzed peripheral blood mononuclear cells from patients with MM during and after SARS-CoV-2 vaccination and breakthrough infection (BTI) using combined whole-transcriptome and surface proteome single-cell profiling with functional serological and T-cell validation in 58 MM patients. Our results demonstrate that vaccine-responders showed a significant overrepresentation of cytotoxic CD4+ T- and mature CD38+ NK-cells expressing FAS+/TIM3+ with a robust cytokine-responsiveness, such as type-I-interferon-, IL-12- and TNF-α-mediated signaling. Patients with MM experiencing BTI developed strong serological and cellular responses and exhibited similar cytokine-responsive immune cell patterns as vaccine-responders. This study can expand our understanding of molecular and cellular patterns associated with immunization responses and may benefit the design of improved vaccination strategies in immunocompromised patients.

NAD+ metabolism is a key modulator of bacterial respiratory epithelial infections

Lower respiratory tract infections caused by Streptococcus pneumoniae (Spn) are a leading cause of death globally. Here we investigate the bronchial epithelial cellular response to Spn infection on a transcriptomic, proteomic and metabolic level. We found the NAD+ salvage pathway to be dysregulated upon infection in a cell line model, primary human lung tissue and in vivo in rodents, leading to a reduced production of NAD+. Knockdown of NAD+ salvage enzymes (NAMPT, NMNAT1) increased bacterial replication. NAD+ treatment of Spn inhibited its growth while growth of other respiratory pathogens improved. Boosting NAD+ production increased NAD+ levels in immortalized and primary cells and decreased bacterial replication upon infection. NAD+ treatment of Spn dysregulated the bacterial metabolism and reduced intrabacterial ATP. Enhancing the bacterial ATP metabolism abolished the antibacterial effect of NAD+. Thus, we identified the NAD+ salvage pathway as an antibacterial pathway in Spn infections, predicting an antibacterial mechanism of NAD+.

Transglutaminase 2 is associated with adverse colorectal cancer survival and represents a therapeutic target

Molecular markers for predicting prognosis of colorectal cancer (CRC) patients are urgently needed for effective disease management. We reported previously that the multifunctional enzyme Transglutaminase 2 (TGM2) is essential for CRC cell survival by inactivation of the tumor suppressor p53. Based on these data, we determined the clinical relevance of TGM2 expression and explored its potential as prognostic marker and therapeutic target in CRC. We profiled TGM2 protein expression in tumor samples of 279 clinically characterized CRC patients using immunohistochemical staining. TGM2 expression was upregulated in matched tumor samples in comparison to normal tissue. A strong TGM2 expression was associated with advanced tumor stages and predicted worse prognosis regarding progression-free and overall-survival, even at early stages. Inhibition of TGM2 in CRC cell lines by the inhibitors LDN27219 and Tyrphostin resulted in a strong reduction of cancer cell proliferation and tumorsphere formation in vitro by induction of p53-mediated apoptosis. Primary patient-derived tumorsphere formation was significantly reduced by inhibition of TGM2. Treatment of mice with TGM2 inhibitors exhibited a significant deceleration of tumor progression. Our data indicate that high TGM2 expression in CRC is associated with worse prognosis and may serve as a therapeutic target in CRC patients with strong TGM2 expression.

Full Profiling of GE81112A, an Underexplored Tetrapeptide Antibiotic with Activity against Gram-Negative Pathogens

After the first total synthesis combined with structure revision, we performed thorough in vitro and in vivo profiling of the underexplored tetrapeptide GE81112A. From the determination of the biological activity spectrum and physicochemical and early absorption-distribution-metabolism-excretion-toxicity (eADMET) properties, as well as in vivo data regarding tolerability and pharmacokinetics (PK) in mice and efficacy in an Escherichia coli-induced septicemia model, we were able to identify the critical and limiting parameters of the original hit compound. Thus, the generated data will serve as the basis for further compound optimization programs and developability assessments to identify candidates for preclinical/clinical development derived from GE81112A as the lead structure. IMPORTANCE The spread of antimicrobial resistance (AMR) is becoming a more and more important global threat to human health. With regard to current medical needs, penetration into the site of infection represents the major challenge in the treatment of infections caused by Gram-positive bacteria. Considering infections associated with Gram-negative bacteria, resistance is a major issue. Obviously, novel scaffolds for the design of new antibacterials in this arena are urgently needed to overcome this crisis. Such a novel potential lead structure is represented by the GE81112 compounds, which inhibit protein synthesis by interacting with the small 30S ribosomal subunit using a binding site distinct from that of other known ribosome-targeting antibiotics. Therefore, the tetrapeptide antibiotic GE81112A was chosen for further exploration as a potential lead for the development of antibiotics with a new mode of action against Gram-negative bacteria.

Machine-guided path sampling to discover mechanisms of molecular self-organization

Molecular self-organization driven by concerted many-body interactions produces the ordered structures that define both inanimate and living matter. Here we present an autonomous path sampling algorithm that integrates deep learning and transition path theory to discover the mechanism of molecular self-organization phenomena. The algorithm uses the outcome of newly initiated trajectories to construct, validate and-if needed-update quantitative mechanistic models. Closing the learning cycle, the models guide the sampling to enhance the sampling of rare assembly events. Symbolic regression condenses the learned mechanism into a human-interpretable form in terms of relevant physical observables. Applied to ion association in solution, gas-hydrate crystal formation, polymer folding and membrane-protein assembly, we capture the many-body solvent motions governing the assembly process, identify the variables of classical nucleation theory, uncover the folding mechanism at different levels of resolution and reveal competing assembly pathways. The mechanistic descriptions are transferable across thermodynamic states and chemical space.

Interaction of dynamic error signals in saccade adaptation

Motor adaptation maintains movement accuracy. To evaluate movement accuracy, motor adaptation relies on an error signal, generated by the movement target, while suppressing error signals from irrelevant objects in the vicinity. Previous work used static testing environments, where all information required to evaluate movement accuracy was available simultaneously. Using saccadic eye movements as a model for motor adaptation, we tested how movement accuracy is maintained in dynamic environments, where the availability of conflicting error signals varied over time. Participants made a vertical saccade toward a target (either a small square or a large ring). Upon saccade detection, two candidate stimuli were shown left and right of the target, and participants were instructed to discriminate a feature on one of the candidates. Critically, candidate stimuli were presented sequentially, and saccade adaptation, thus, had to resolve a conflict between a task-relevant and a task-irrelevant error signal that were separated in space and time. We found that the saccade target influenced several aspects of oculomotor learning. In presence of a small target, saccade adaptation evaluated movement accuracy based on the first available error signal after the saccade, irrespective of its task relevance. However, a large target not only allowed for greater flexibility when evaluating movement accuracy, but it also promoted a stronger contribution of strategic behavior when compensating inaccurate saccades. Our results demonstrate how motor adaptation maintains movement accuracy in dynamic environments, and how properties of the visual environment modulate the relative contribution of different learning processes.NEW & NOTEWORTHY Motor adaptation is typically studied in static environments, where all information that is required to evaluate movement accuracy is available simultaneously. Here, using saccadic eye movements as a model, we studied motor adaptation in a dynamic environment, where the availability of conflicting information about movement accuracy varied over time. We demonstrate that properties of the visual environment determine how dynamic movement errors are corrected.

Transglutaminase 2 promotes tumorigenicity of colon cancer cells by inactivation of the tumor suppressor p53

Despite a high clinical need for the treatment of colorectal carcinoma (CRC) as the second leading cause of cancer-related deaths, targeted therapies are still limited. The multifunctional enzyme Transglutaminase 2 (TGM2), which harbors transamidation and GTPase activity, has been implicated in the development and progression of different types of human cancers. However, the mechanism and role of TGM2 in colorectal cancer are poorly understood. Here, we present TGM2 as a promising drug target.In primary patient material of CRC patients, we detected an increased expression and enzymatic activity of TGM2 in colon cancer tissue in comparison to matched normal colon mucosa cells. The genetic ablation of TGM2 in CRC cell lines using shRNAs or CRISPR/Cas9 inhibited cell expansion and tumorsphere formation. In vivo, tumor initiation and growth were reduced upon genetic knockdown of TGM2 in xenotransplantations. TGM2 ablation led to the induction of Caspase-3-driven apoptosis in CRC cells. Functional rescue experiments with TGM2 variants revealed that the transamidation activity is critical for the pro-survival function of TGM2. Transcriptomic and protein-protein interaction analyses applying various methods including super-resolution and time-lapse microscopy showed that TGM2 directly binds to the tumor suppressor p53, leading to its inactivation and escape of apoptosis induction.We demonstrate here that TGM2 is an essential survival factor in CRC, highlighting the therapeutic potential of TGM2 inhibitors in CRC patients with high TGM2 expression. The inactivation of p53 by TGM2 binding indicates a general anti-apoptotic function, which may be relevant in cancers beyond CRC.

Identification of Fungal Limonene-3-Hydroxylase for Biotechnological Menthol Production

More than 30,000 tons of menthol are produced every year as a flavor and fragrance compound or as a medical component. So far, only extraction from plant material and chemical synthesis are possible. An alternative approach for menthol production could be a biotechnological-chemical process with ideally only two conversion steps, starting from (+)-limonene, which is a side product of the citrus processing industry. The first step requires a limonene-3-hydroxylase (L3H) activity that specifically catalyzes hydroxylation of limonene at carbon atom 3. Several protein engineering strategies have already attempted to create limonene-3-hydroxylases from bacterial cytochrome P450 monooxygenases (CYPs, or P450s), which can be efficiently expressed in bacterial hosts. However, their regiospecificity is rather low compared to that of the highly selective L3H enzymes from the biosynthetic pathway for menthol in Mentha species. The only naturally occurring limonene-3-hydroxylase activity identified in microorganisms so far was reported for a strain of the black yeast-like fungus Hormonema sp. in South Africa. We have discovered additional fungi that can catalyze the intended reaction and identified potential CYP-encoding genes within the genome sequence of one of the strains. Using heterologous gene expression and biotransformation experiments in yeasts, we were able to identify limonene-3-hydroxylases from Aureobasidium pullulans and Hormonema carpetanum Further characterization of the A. pullulans enzyme demonstrated its high stereospecificity and regioselectivity, its potential for limonene-based menthol production, and its additional ability to convert α- and β-pinene to verbenol and pinocarveol, respectively.IMPORTANCE (-)-Menthol is an important flavor and fragrance compound and furthermore has medicinal uses. To realize a two-step synthesis starting from renewable (+)-limonene, a regioselective limonene-3-hydroxylase enzyme is necessary. We identified enzymes from two different fungi which catalyze this hydroxylation reaction and represent an important module for the development of a biotechnological process for (-)-menthol production from renewable (+)-limonene.

gammaBOriS: Identification and Taxonomic Classification of Origins of Replication in Gammaproteobacteria using Motif-based Machine Learning

The biology of bacterial cells is, in general, based on information encoded on circular chromosomes. Regulation of chromosome replication is an essential process that mostly takes place at the origin of replication (oriC), a locus unique per chromosome. Identification of high numbers of oriC is a prerequisite for systematic studies that could lead to insights into oriC functioning as well as the identification of novel drug targets for antibiotic development. Current methods for identifying oriC sequences rely on chromosome-wide nucleotide disparities and are therefore limited to fully sequenced genomes, leaving a large number of genomic fragments unstudied. Here, we present gammaBOriS (Gammaproteobacterial oriC Searcher), which identifies oriC sequences on gammaproteobacterial chromosomal fragments. It does so by employing motif-based machine learning methods. Using gammaBOriS, we created BOriS DB, which currently contains 25,827 gammaproteobacterial oriC sequences from 1,217 species, thus making it the largest available database for oriC sequences to date. Furthermore, we present gammaBOriTax, a machine-learning based approach for taxonomic classification of oriC sequences, which was trained on the sequences in BOriS DB. Finally, we extracted the motifs relevant for identification and classification decisions of the models. Our results suggest that machine learning sequence classification approaches can offer great support in functional motif identification.