Search

BackgroundMachine learning models require large datasets that may be siloed across different health care institutions. Machine learning studies that focus on COVID-19 have been limited to single-hospital data, which limits model generalizability. ObjectiveWe aimed to use federated learning, a machine learning technique that avoids locally aggregating raw clinical data across multiple institutions, to predict mortality in hospitalized patients with COVID-19 within 7 days. MethodsPatient data were collected from the electronic health records of 5 hospitals within the Mount Sinai Health System. Logistic regression with L1 regularization/least absolute shrinkage and selection operator (LASSO) and multilayer perceptron (MLP) models were trained by using local data at each site. We developed a pooled model with combined data from all 5 sites, and a federated model that only shared parameters with a central aggregator. ResultsThe LASSOfederated model outperformed the LASSOlocal model at 3 hospitals, and the MLPfederated model performed better than the MLPlocal model at all 5 hospitals, as determined by the area under the receiver operating characteristic curve. The LASSOpooled model outperformed the LASSOfederated model at all hospitals, and the MLPfederated model outperformed the MLPpooled model at 2 hospitals. ConclusionsThe federated learning of COVID-19 electronic health record data shows promise in developing robust predictive models without compromising patient privacy.

Alzheimer's Disease (AD) is a neurodegenerative disorder that is still not fully understood. Sex modifies AD vulnerability, but the reasons for this are largely unknown. We utilize two independent electronic medical record (EMR) systems across 44,288 patients to perform deep clinical phenotyping and network analysis to gain insight into clinical characteristics and sex-specific clinical associations in AD. Embeddings and network representation of patient diagnoses demonstrate greater comorbidity interactions in AD in comparison to matched controls. Enrichment analysis identifies multiple known and new diagnostic, medication, and lab result associations across the whole cohort and in a sex-stratified analysis. With this data-driven method of phenotyping, we can represent AD complexity and generate hypotheses of clinical factors that can be followed-up for further diagnostic and predictive analyses, mechanistic understanding, or drug repurposing and therapeutic approaches.

The evident genetic, pathological, and clinical heterogeneity of Alzheimer's disease (AD) poses challenges for traditional drug development. We conducted a computational drug repurposing screen for drugs to treat apolipoprotein (apo) E4-related AD. We first established apoE-genotype-dependent transcriptomic signatures of AD by analyzing publicly-available human brain database. We then queried these signatures against the Connectivity Map database containing transcriptomic perturbations of >1300 drugs to identify those that best reverse apoE-genotype-specific AD signatures. Bumetanide was identified as a top drug for apoE4 AD. Bumetanide treatment of apoE4 mice without or with Aβ accumulation rescued electrophysiological, pathological, or cognitive deficits. Single-nucleus RNA-sequencing revealed transcriptomic reversal of AD signatures in specific cell types in these mice, a finding confirmed in apoE4-iPSC-derived neurons. In humans, bumetanide exposure was associated with a significantly lower AD prevalence in individuals over the age of 65 in two electronic health record databases, suggesting effectiveness of bumetanide in preventing AD.

Hu11B6 is a monoclonal antibody that internalizes in cells expressing androgen receptor (AR)-regulated prostate-specific enzyme human kallikrein-related peptidase 2 (hK2; KLK2). In multiple rodent models, Actinium-225-labeled hu11B6-IgG1 ([225Ac]hu11B6-IgG1) has shown promising treatment efficacy. In the present study, we investigated options to enhance and optimize [225Ac]hu11B6 treatment. First, we evaluated the possibility of exploiting IgG3, the IgG subclass with superior activation of complement and ability to mediate FC-γ-receptor binding, for immunotherapeutically enhanced hK2 targeted α-radioimmunotherapy. Second, we compared the therapeutic efficacy of a single high activity vs. fractionated activity. Finally, we used RNA sequencing to analyze the genomic signatures of prostate cancer that progressed after targeted α-therapy. [225Ac]hu11B6-IgG3 was a functionally enhanced alternative to [225Ac]hu11B6-IgG1 but offered no improvement of therapeutic efficacy. Progression-free survival was slightly increased with a single high activity compared to fractionated activity. Tumor-free animals succumbing after treatment revealed no evidence of treatment-associated toxicity. In addition to up-regulation of canonical aggressive prostate cancer genes, such as MMP7, ETV1, NTS, and SCHLAP1, we also noted a significant decrease in both KLK3 (prostate-specific antigen ) and FOLH1 (prostate-specific membrane antigen) but not in AR and KLK2, demonstrating efficacy of sequential [225Ac]hu11B6 in a mouse model.

Ocean Sampling Day was initiated by the EU-funded Micro B3 (Marine Microbial Biodiversity, Bioinformatics, Biotechnology) project to obtain a snapshot of the marine microbial biodiversity and function of the world's oceans. It is a simultaneous global mega-sequencing campaign aiming to generate the largest standardized microbial data set in a single day. This will be achievable only through the coordinated efforts of an Ocean Sampling Day Consortium, supportive partnerships and networks between sites. This commentary outlines the establishment, function and aims of the Consortium and describes our vision for a sustainable study of marine microbial communities and their embedded functional traits.

The co-authors of this paper hereby state their intention to work together to launch the Genomic Observatories Network (GOs Network) for which this document will serve as its Founding Charter. We define a Genomic Observatory as an ecosystem and/or site subject to long-term scientific research, including (but not limited to) the sustained study of genomic biodiversity from single-celled microbes to multicellular organisms.An international group of 64 scientists first published the call for a global network of Genomic Observatories in January 2012. The vision for such a network was expanded in a subsequent paper and developed over a series of meetings in Bremen (Germany), Shenzhen (China), Moorea (French Polynesia), Oxford (UK), Pacific Grove (California, USA), Washington (DC, USA), and London (UK). While this community-building process continues, here we express our mutual intent to establish the GOs Network formally, and to describe our shared vision for its future. The views expressed here are ours alone as individual scientists, and do not necessarily represent those of the institutions with which we are affiliated.

AbstractWe are entering a new era in genomics–that of large-scale, place-based, highly contextualized genomic research. Here we review this emerging paradigm shift and suggest that sites of utmost scientific importance be expanded into ‘Genomic Observatories’ (GOs). Investment in GOs should focus on the digital characterization of whole ecosystems, from all-taxa biotic inventories to time-series ’omics studies. The foundational layer of biodiversity–genetic variation–would thus be mainstreamed into Earth Observation systems enabling predictive modelling of biodiversity dynamics and resultant impacts on ecosystem services.

FLAME allows complex models to be automatically parallelised on High Performance Computing (HPC) grids enabling large number of agents to be simulated over short periods of time. Modellers are hindered by complexities of porting models on parallel platforms and time taken to run large simulations on a single machine, which FLAME overcomes. Three case studies from different disciplines were modelled using FLAME, and are presented along with their performance results on a grid.

Abstract Background A steep drop in the cost of next-generation sequencing during recent years has made the technology affordable to the majority of researchers, but downstream bioinformatic analysis still poses a resource bottleneck for smaller laboratories and institutes that do not have access to substantial computational resources. Sequencing instruments are typically bundled with only the minimal processing and storage capacity required for data capture during sequencing runs. Given the scale of sequence datasets, scientific value cannot be obtained from acquiring a sequencer unless it is accompanied by an equal investment in informatics infrastructure. Results Cloud BioLinux is a publicly accessible Virtual Machine (VM) that enables scientists to quickly provision on-demand infrastructures for high-performance bioinformatics computing using cloud platforms. Users have instant access to a range of pre-configured command line and graphical software applications, including a full-featured desktop interface, documentation and over 135 bioinformatics packages for applications including sequence alignment, clustering, assembly, display, editing, and phylogeny. Each tool's functionality is fully described in the documentation directly accessible from the graphical interface of the VM. Besides the Amazon EC2 cloud, we have started instances of Cloud BioLinux on a private Eucalyptus cloud installed at the J. Craig Venter Institute, and demonstrated access to the bioinformatic tools interface through a remote connection to EC2 instances from a local desktop computer. Documentation for using Cloud BioLinux on EC2 is available from our project website, while a Eucalyptus cloud image and VirtualBox Appliance is also publicly available for download and use by researchers with access to private clouds. Conclusions Cloud BioLinux provides a platform for developing bioinformatics infrastructures on the cloud. An automated and configurable process builds Virtual Machines, allowing the development of highly customized versions from a shared code base. This shared community toolkit enables application specific analysis platforms on the cloud by minimizing the effort required to prepare and maintain them.

Tauopathies are a heterogeneous group of neurodegenerative disorders that are clinically and pathologically distinct from Alzheimer's disease (AD) having tau inclusions in neurons and/or glia as their most prominent neuropathological feature. BCI-838 (MGS00210) is a group II metabotropic glutamate receptor (mGluR2/3) antagonist pro-drug. Previously, we reported that orally administered BCI-838 improved learning behavior and reduced anxiety in Dutch (APP E693Q ) transgenic mice, a model of the pathological accumulation of Aβ oligomers found in AD. Herein, we investigated effects of BCI-838 on PS19 male mice that express the tauopathy mutation MAPT P301S associated with human frontotemporal lobar degeneration (FTLD). These mice develop an aging-related tauopathy without amyloid accumulation. Mice were divided into three experimental groups: (1) non-transgenic wild type mice treated with vehicle, (2) PS19 mice treated with vehicle and (3) PS19 mice treated with 5 mg/kg BCI-838. Groups of 10-13 mice were utilized. Vehicle or BCI-838 was administered by oral gavage for 4 weeks. Behavioral testing consisting of a novel object recognition task was conducted after drug administration. Two studies were performed beginning treatment of mice at 3 or 7 months of age. One month of BCI-838 treatment rescued deficits in recognition memory in PS19 mice whether treatment was begun at 3 or 7 months of age. These studies extend the potential utility of BCI-838 to neurodegenerative conditions that have tauopathy as their underlying basis. They also suggest an mGluR2/3 dependent mechanism as a basis for the behavioral deficits in PS19 mice., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Published by Elsevier B.V.)