The FeatureCloud Platform for Federated Learning in Biomedicine: Unified Approach.

BACKGROUND: Machine learning and artificial intelligence have shown promising results in many areas and are driven by the increasing amount of available data. However, these data are often distributed across different institutions and cannot be easily shared owing to strict privacy regulations. Federated learning (FL) allows the training of distributed machine learning models without sharing sensitive data. In addition, the implementation is time-consuming and requires advanced programming skills and complex technical infrastructures. OBJECTIVE: Various tools and frameworks have been developed to simplify the development of FL algorithms and provide the necessary technical infrastructure. Although there are many high-quality frameworks, most focus only on a single application case or method. To our knowledge, there are no generic frameworks, meaning that the existing solutions are restricted to a particular type of algorithm or application field. Furthermore, most of these frameworks provide an application programming interface that needs programming knowledge. There is no collection of ready-to-use FL algorithms that are extendable and allow users (eg, researchers) without programming knowledge to apply FL. A central FL platform for both FL algorithm developers and users does not exist. This study aimed to address this gap and make FL available to everyone by developing FeatureCloud, an all-in-one platform for FL in biomedicine and beyond. METHODS: The FeatureCloud platform consists of 3 main components: a global frontend, a global backend, and a local controller. Our platform uses a Docker to separate the local acting components of the platform from the sensitive data systems. We evaluated our platform using 4 different algorithms on 5 data sets for both accuracy and runtime. RESULTS: FeatureCloud removes the complexity of distributed systems for developers and end users by providing a comprehensive platform for executing multi-institutional FL analyses and implementing FL algorithms. Through its integrated artificial intelligence store, federated algorithms can easily be published and reused by the community. To secure sensitive raw data, FeatureCloud supports privacy-enhancing technologies to secure the shared local models and assures high standards in data privacy to comply with the strict General Data Protection Regulation. Our evaluation shows that applications developed in FeatureCloud can produce highly similar results compared with centralized approaches and scale well for an increasing number of participating sites. CONCLUSIONS: FeatureCloud provides a ready-to-use platform that integrates the development and execution of FL algorithms while reducing the complexity to a minimum and removing the hurdles of federated infrastructure. Thus, we believe that it has the potential to greatly increase the accessibility of privacy-preserving and distributed data analyses in biomedicine and beyond.

Human gestational N-methyl-d-aspartate receptor autoantibodies impair neonatal murine brain function.

OBJECTIVE: Maternal autoantibodies are a risk factor for impaired brain development in offspring. Antibodies (ABs) against the NR1 (GluN1) subunit of the N-methyl-d-aspartate receptor (NMDAR) are among the most frequently diagnosed anti-neuronal surface ABs, yet little is known about effects on fetal development during pregnancy. METHODS: We established a murine model of in utero exposure to human recombinant NR1 and isotype-matched nonreactive control ABs. Pregnant C57BL/6J mice were intraperitoneally injected on embryonic days 13 and 17 each with 240µg of human monoclonal ABs. Offspring were investigated for acute and chronic effects on NMDAR function, brain development, and behavior. RESULTS: Transferred NR1 ABs enriched in the fetus and bound to synaptic structures in the fetal brain. Density of NMDAR was considerably reduced (up to -49.2%) and electrophysiological properties were altered, reflected by decreased amplitudes of spontaneous excitatory postsynaptic currents in young neonates (-34.4%). NR1 AB-treated animals displayed increased early postnatal mortality (+27.2%), impaired neurodevelopmental reflexes, altered blood pH, and reduced bodyweight. During adolescence and adulthood, animals showed hyperactivity (+27.8% median activity over 14 days), lower anxiety, and impaired sensorimotor gating. NR1 ABs caused long-lasting neuropathological effects also in aged mice (10 months), such as reduced volumes of cerebellum, midbrain, and brainstem. INTERPRETATION: The data collectively support a model in which asymptomatic mothers can harbor low-level pathogenic human NR1 ABs that are diaplacentally transferred, causing neurotoxic effects on neonatal development. Thus, AB-mediated network changes may represent a potentially treatable neurodevelopmental congenital brain disorder contributing to lifelong neuropsychiatric morbidity in affected children. ANN NEUROL 2019;86:656-670.

Wnt signaling positively regulates endothelial cell fate specification in the Fli1a-positive progenitor population via Lef1.

During vertebrate embryogenesis, vascular endothelial cells (ECs) and primitive erythrocytes become specified within close proximity in the posterior lateral plate mesoderm (LPM) from a common progenitor. However, the signaling cascades regulating the specification into either lineage remain largely elusive. Here, we analyze the contribution of ß-catenin dependent Wnt signaling to EC and erythrocyte specification during zebrafish embryogenesis. We generated novel ß-catenin dependent Wnt signaling reporters which, by using destabilized fluorophores (Venus-Pest, dGFP), specifically allow us to detect Wnt signaling responses in narrow time windows as well as in spatially restricted domains, defined by Cre recombinase expression (Tg(axin2BAC:Venus-Pest)mu288; Tg(14TCF:loxP-STOP-loxP-dGFP)mu202). We therefore can detect ß-catenin dependent Wnt signaling activity in a subset of the Fli1a-positive progenitor population. Additionally, we show that mesodermal Wnt3a-mediated signaling via the transcription factor Lef1 positively regulates EC specification (defined by kdrl expression) at the expense of primitive erythrocyte specification (defined by gata1 expression) in zebrafish embryos. Using mesoderm derived from human embryonic stem cells, we identified the same principle of Wnt signaling dependent EC specification in conjunction with auto-upregulation of LEF1. Our data indicate a novel role of ß-catenin dependent Wnt signaling in regulating EC specification during vasculogenesis.

Human cerebrospinal fluid monoclonal N-methyl-D-aspartate receptor autoantibodies are sufficient for encephalitis pathogenesis.

SEE ZEKERIDOU AND LENNON DOI101093/AWW213 FOR A SCIENTIFIC COMMENTARY ON THIS ARTICLE: Anti-N-methyl-D-aspartate receptor (NMDAR) encephalitis is a recently discovered autoimmune syndrome associated with psychosis, dyskinesias, and seizures. Little is known about the cerebrospinal fluid autoantibody repertoire. Antibodies against the NR1 subunit of the NMDAR are thought to be pathogenic; however, direct proof is lacking as previous experiments could not distinguish the contribution of further anti-neuronal antibodies. Using single cell cloning of full-length immunoglobulin heavy and light chain genes, we generated a panel of recombinant monoclonal NR1 antibodies from cerebrospinal fluid memory B cells and antibody secreting cells of NMDAR encephalitis patients. Cells typically carried somatically mutated immunoglobulin genes and had undergone class-switching to immunoglobulin G, clonally expanded cells carried identical somatic hypermutation patterns. A fraction of NR1 antibodies were non-mutated, thus resembling 'naturally occurring antibodies' and indicating that tolerance induction against NMDAR was incomplete and somatic hypermutation not essential for functional antibodies. However, only a small percentage of cerebrospinal fluid-derived antibodies reacted against NR1. Instead, nearly all further antibodies bound specifically to diverse brain-expressed epitopes including neuronal surfaces, suggesting that a broad repertoire of antibody-secreting cells enrich in the central nervous system during encephalitis. Our functional data using primary hippocampal neurons indicate that human cerebrospinal fluid-derived monoclonal NR1 antibodies alone are sufficient to cause neuronal surface receptor downregulation and subsequent impairment of NMDAR-mediated currents, thus providing ultimate proof of antibody pathogenicity. The observed formation of immunological memory might be relevant for clinical relapses.

Flimma: a federated and privacy-aware tool for differential gene expression analysis.

Aggregating transcriptomics data across hospitals can increase sensitivity and robustness of differential expression analyses, yielding deeper clinical insights. As data exchange is often restricted by privacy legislation, meta-analyses are frequently employed to pool local results. However, the accuracy might drop if class labels are inhomogeneously distributed among cohorts. Flimma ( https://exbio.wzw.tum.de/flimma/ ) addresses this issue by implementing the state-of-the-art workflow limma voom in a federated manner, i.e., patient data never leaves its source site. Flimma results are identical to those generated by limma voom on aggregated datasets even in imbalanced scenarios where meta-analysis approaches fail.

Exploring the SARS-CoV-2 virus-host-drug interactome for drug repurposing.

Coronavirus Disease-2019 (COVID-19) is an infectious disease caused by the SARS-CoV-2 virus. Various studies exist about the molecular mechanisms of viral infection. However, such information is spread across many publications and it is very time-consuming to integrate, and exploit. We develop CoVex, an interactive online platform for SARS-CoV-2 host interactome exploration and drug (target) identification. CoVex integrates virus-human protein interactions, human protein-protein interactions, and drug-target interactions. It allows visual exploration of the virus-host interactome and implements systems medicine algorithms for network-based prediction of drug candidates. Thus, CoVex is a resource to understand molecular mechanisms of pathogenicity and to prioritize candidate therapeutics. We investigate recent hypotheses on a systems biology level to explore mechanistic virus life cycle drivers, and to extract drug repurposing candidates. CoVex renders COVID-19 drug research systems-medicine-ready by giving the scientific community direct access to network medicine algorithms. It is available at https://exbio.wzw.tum.de/covex/.