Search

Detecting epistatic drivers of human phenotypes is a considerable challenge. Traditional approaches use regression to sequentially test multiplicative interaction terms involving pairs of genetic variants. For higher-order interactions and genome-wide large-scale data, this strategy is computationally intractable. Moreover, multiplicative terms used in regression modeling may not capture the form of biological interactions. Building on the Predictability, Computability, Stability (PCS) framework, we introduce the epiTree pipeline to extract higher-order interactions from genomic data using tree-based models. The epiTree pipeline first selects a set of variants derived from tissue-specific estimates of gene expression. Next, it uses iterative random forests (iRF) to search training data for candidate Boolean interactions (pairwise and higher-order). We derive significance tests for interactions, based on a stabilized likelihood ratio test, by simulating Boolean tree-structured null (no epistasis) and alternative (epistasis) distributions on hold-out test data. Finally, our pipeline computes PCS epistasis p-values that probabilisticly quantify improvement in prediction accuracy via bootstrap sampling on the test set. We validate the epiTree pipeline in two case studies using data from the UK Biobank: predicting red hair and multiple sclerosis (MS). In the case of predicting red hair, epiTree recovers known epistatic interactions surrounding MC1R and novel interactions, representing non-linearities not captured by logistic regression models. In the case of predicting MS, a more complex phenotype than red hair, epiTree rankings prioritize novel interactions surrounding HLA-DRB1, a variant previously associated with MS in several populations. Taken together, these results highlight the potential for epiTree rankings to help reduce the design space for follow up experiments.

While domain-specific data augmentation can be useful in training neural networks for medical imaging tasks, such techniques have not been widely used to date. Our objective was to test whether domain-specific data augmentation is useful for medical imaging using a well-benchmarked task: view classification on fetal ultrasound FETAL-125 and OB-125 datasets. We found that using a context-preserving cut-paste strategy, we could create valid training data as measured by performance of the resulting trained model on the benchmark test dataset. When used in an online fashion, models trained on this hybrid data performed similarly to those trained using traditional data augmentation (FETAL-125 F-score 85.33 ± 0.24 vs 86.89 ± 0.60, p-value 0.014; OB-125 F-score 74.60 ± 0.11 vs 72.43 ± 0.62, p-value 0.004). Furthermore, the ability to perform augmentations during training time, as well as the ability to apply chosen augmentations equally across data classes, are important considerations in designing a bespoke data augmentation. Finally, we provide open-source code to facilitate running bespoke data augmentations in an online fashion. Taken together, this work expands the ability to design and apply domain-guided data augmentations for medical imaging tasks.

Detecting epistatic drivers of human phenotypes is a considerable challenge. Traditional approaches use regression to sequentially test multiplicative interaction terms involving pairs of genetic variants. For higher-order interactions and genome-wide large-scale data, this strategy is computationally intractable. Moreover, multiplicative terms used in regression modeling may not capture the form of biological interactions. Building on the Predictability, Computability, Stability (PCS) framework, we introduce the epiTree pipeline to extract higher-order interactions from genomic data using tree-based models. The epiTree pipeline first selects a set of variants derived from tissue-specific estimates of gene expression. Next, it uses iterative random forests (iRF) to search training data for candidate Boolean interactions (pairwise and higher-order). We derive significance tests for interactions, based on a stabilized likelihood ratio test, by simulating Boolean tree-structured null (no epistasis) and alternative (epistasis) distributions on hold-out test data. Finally, our pipeline computes PCS epistasis p-values that probabilisticly quantify improvement in prediction accuracy via bootstrap sampling on the test set. We validate the epiTree pipeline in two case studies using data from the UK Biobank: predicting red hair and multiple sclerosis (MS). In the case of predicting red hair, epiTree recovers known epistatic interactions surrounding MC1R and novel interactions, representing non-linearities not captured by logistic regression models. In the case of predicting MS, a more complex phenotype than red hair, epiTree rankings prioritize novel interactions surrounding HLA-DRB1, a variant previously associated with MS in several populations. Taken together, these results highlight the potential for epiTree rankings to help reduce the design space for follow up experiments.

With over 1,400 species worldwide, bats represent the second largest order of mammals after rodents, and are known to host major zoonotic pathogens. Here, we estimate the presence of pathogens in autochthonous bat populations. First, we set out to check our samples for PCR amplification efficiency by assessing the occurrence of inhibited PCR reactions from different types of bat samples with amplifying the housekeeping gene β-actin. Second, we investigated the presence of five targeted pathogens in a French bat population using PCR. We targeted viral RNA of Canine distemper virus, Alphacoronavirus, Lyssavirus, Rotavirus and bacterial Leptospira DNA. To do so, we screened for these viruses in bat faecal samples as well as in oropharyngeal swab samples. The presence of Leptospira was assessed in urine, kidney, lung and faecal samples. Results showed a frequency of inhibited reactions ranging from 5 to 60% of samples, varying according to the sample itself and also suspected to vary according to sampling method and the storage buffer solution used, demonstrating the importance of the sampling and storage on the probability of obtaining negative PCR results. For pathogen assessment, rotavirus and alphacoronavirus RNA were detected in Myotis myotis, Myotis daubentonii, Myotis emarginatus and Rhinolophus ferrumequinum bats. Rotaviruses were also detected in Barbastella barbastellus. The presence of alphacoronavirus also varied seasonally, with higher frequencies in late summer and October, suggesting that juveniles potentially play an important role in the dynamics of these viruses. Leptospira DNA was detected in M. myotis and M. daubentonii colonies. The 16S rRNA sequences obtained from Leptospira positive samples showed 100% genetic identity with L. borgpetersenii. Neither canine distemper virus nor lyssavirus RNA were detected in any of the tested samples. This study is the first to show the presence of Leptospira in autochthonous French bats in addition to coronavirus and rotavirus RNA previously reported in European autochthonous bats.

While domain-specific data augmentation can be useful in training neural networks for medical imaging tasks, such techniques have not been widely used to date. Our objective was to test whether domain-specific data augmentation is useful for medical imaging using a well-benchmarked task: view classification on fetal ultrasound FETAL-125 and OB-125 datasets. We found that using a context-preserving cut-paste strategy, we could create valid training data as measured by performance of the resulting trained model on the benchmark test dataset. When used in an online fashion, models trained on this hybrid data performed similarly to those trained using traditional data augmentation (FETAL-125 F-score 85.33 ± 0.24 vs 86.89 ± 0.60, p-value 0.014; OB-125 F-score 74.60 ± 0.11 vs 72.43 ± 0.62, p-value 0.004). Furthermore, the ability to perform augmentations during training time, as well as the ability to apply chosen augmentations equally across data classes, are important considerations in designing a bespoke data augmentation. Finally, we provide open-source code to facilitate running bespoke data augmentations in an online fashion. Taken together, this work expands the ability to design and apply domain-guided data augmentations for medical imaging tasks.

For most of the COVID-19 pandemic, the daily focus has been on the number of cases, and secondarily, deaths. The most recent wave was caused by the omicron variant, first identified at the end of 2021 and the dominant variant through the first part of 2022. South Africa, one of the first countries to experience and report data regarding omicron (variant 21.K), reported far fewer deaths, even as the number of reported cases rapidly eclipsed previous peaks. However, as the omicron wave has progressed, time series show that it has been markedly different from prior waves. To more readily visualize the dynamics of cases and deaths, it is natural to plot deaths per million against cases per million. Unlike the time-series plots of cases or deaths that have become daily features of pandemic updates during the pandemic, which have time as the x-axis, in a plot of deaths vs. cases, time is implicit, and is indicated in relation to the starting point. Here we present and briefly examine such plots from a number of countries and from the world as a whole, illustrating how they summarize features of the pandemic in ways that illustrate how, in most places, the omicron wave is very different from those that came before. Code for generating these plots for any country is provided in an automatically updating GitHub repository.

With more than 1400 chiropteran species identified to date, bats comprise one-fifth of all mammalian species worldwide. Many studies have associated viral zoonoses with 45 different species of bats in the EU, which cluster within 5 families of bats. For example, the Serotine bats are infected by European Bat 1 Lyssavirus throughout Europe while Myotis bats are shown infected by coronavirus, herpesvirus and paramyxovirus. Correct host species identification is important to increase our knowledge of the ecology and evolutionary pattern of bat viruses in the EU. Bat species identification is commonly determined using morphological keys. Morphological determination of bat species from bat carcasses can be limited in some cases, due to the state of decomposition or nearly indistinguishable morphological features in juvenile bats and can lead to misidentifications. The overall objective of our study was to identify insectivorous bat species using molecular biology tools with the amplification of the partial cytochrome b gene of mitochondrial DNA. Two types of samples were tested in this study, bat wing punches and bat faeces. A total of 163 bat wing punches representing 22 species, and 31 faecal pellets representing 7 species were included in the study. From the 163 bat wing punches tested, a total of 159 were genetically identified from amplification of the partial cyt b gene. All 31 faecal pellets were genetically identified based on the cyt b gene. A comparison between morphological and genetic determination showed 21 misidentifications from the 163 wing punches, representing ~12.5% of misidentifications of morphological determination compared with the genetic method, across 11 species. In addition, genetic determination allowed the identification of 24 out of 25 morphologically non-determined bat samples. Our findings demonstrate the importance of a genetic approach as an efficient and reliable method to identify bat species precisely.

BACKGROUND:Health care workers (HCWs) are essential for the delivery of health care services in conflict areas and in rebuilding health systems post-conflict. OBJECTIVE:The aim of this study was to systematically identify and map the published evidence on HCWs in conflict and post-conflict settings. Our ultimate aim is to inform researchers and funders on research gap on this subject and support relevant stakeholders by providing them with a comprehensive resource of evidence about HCWs in conflict and post-conflict settings on a global scale. METHODS:We conducted a systematic mapping of the literature. We included a wide range of study designs, addressing any type of personnel providing health services in either conflict or post-conflict settings. We conducted a descriptive analysis of the general characteristics of the included papers and built two interactive systematic maps organized by country, study design and theme. RESULTS:Out of 13,863 identified citations, we included a total of 474 studies: 304 on conflict settings, 149 on post-conflict settings, and 21 on both conflict and post-conflict settings. For conflict settings, the most studied counties were Iraq (15%), Syria (15%), Israel (10%), and the State of Palestine (9%). The most common types of publication were opinion pieces in conflict settings (39%), and primary studies (33%) in post-conflict settings. In addition, most of the first and corresponding authors were affiliated with countries different from the country focus of the paper. Violence against health workers was the most tackled theme of papers reporting on conflict settings, while workforce performance was the most addressed theme by papers reporting on post-conflict settings. The majority of papers in both conflict and post-conflict settings did not report funding sources (81% and 53%) or conflicts of interest of authors (73% and 62%), and around half of primary studies did not report on ethical approvals (45% and 41%). CONCLUSIONS:This systematic mapping provides a comprehensive database of evidence about HCWs in conflict and post-conflict settings on a global scale that is often needed to inform policies and strategies on effective workforce planning and management and in reducing emigration. It can also be used to identify evidence for policy-relevant questions, knowledge gaps to direct future primary research, and knowledge clusters.

Analysis of the nursing registration databases is a highly informative approach that provides accurate and reliable information supporting evidence based decisions relevant to the nursing workforce planning, management and development. This study presents the first systematic analysis of the nursing registration database in Lebanon. It Reports on the workforce distribution and trends using an updated version of the Order of Nurses in Lebanon (ONL) databases.This study presents a secondary data analysis of a de-identified subset of the updated ONL registration database. The workforce participation status of ONL registered nurses was categorized as active and eligible. For active nurses sectors and sub-sectors of employment were defined. Eligible nurses were categorized as unemployed, working outside nursing and working abroad. SPSS was used to conduct descriptive analysis to present workforce trends of Lebanese nurses for year 2009-2014 as frequencies, percentages and percentage changes.Increases in the size of the Active (35%) and Eligible (86%) nurses were observed over the past six years. The majority of nurses fell in the below 35 years age group (60% in 2014). The hospital sector remained the principle employer, with 87% of Lebanese nurses working in hospitals in 2014. A 173% increases was reported for nurses working abroad.Despite the growth of the Active nursing workforce, the skewed distribution of nurses in the below 35 age group and the growth in the Eligible category, especially for nurses living abroad, raise concerns on the longevity of nurses in the profession and the reasons for their attrition from the workforce.There is a need to investigate the push and pull factors that are affecting nurses and the design of policies and interventions that would encourage nurses to remain active in Lebanon. Furthermore, policies and interventions that would create employment opportunities outside hospitals, especially in the Community sector, are recommended.

In the kidney, the renal proximal tubule (PT) reabsorbs solutes into the peritubular capillaries through active transport. Here, we replicate this reabsorptive function in vitro by engineering a microfluidic PT. The microfluidic PT architecture comprises a porous membrane with user-defined submicron surface topography separating two microchannels representing a PT filtrate lumen and a peritubular capillary lumen. Human PT epithelial cells and microvascular endothelial cells in respective microchannels created a PT-like reabsorptive barrier. Co-culturing epithelial and endothelial cells in the microfluidic architecture enhanced viability, metabolic activity, and compactness of the epithelial layer. The resulting tissue expressed tight junctions, kidney-specific morphology, and polarized expression of kidney markers. The microfluidic PT actively performed sodium-coupled glucose transport, which could be modulated by administration of a sodium-transport inhibiting drug. The microfluidic PT reproduces human physiology at the cellular and tissue levels, and measurable tissue function which can quantify kidney pharmaceutical efficacy and toxicity.

Acute kidney injury, often caused by an ischemic insult, is associated with significant short-term morbidity and mortality, and increased risk of chronic kidney disease. The factors affecting the renal response to injury following ischemia and reperfusion remain to be clarified. We found that the Stem cell antigen-1 (Sca-1), commonly used as a stem cell marker, is heavily expressed in renal tubules of the adult mouse kidney. We evaluated its potential role in the kidney using Sca-1 knockout mice submitted to acute ischemia reperfusion injury (IRI), as well as cultured renal proximal tubular cells in which Sca-1 was stably silenced with shRNA. IRI induced more severe injury in Sca-1 null kidneys, as assessed by increased expression of Kim-1 and Ngal, rise in serum creatinine, abnormal pathology, and increased apoptosis of tubular epithelium, and persistent significant renal injury at day 7 post IRI, when recovery of renal function in control animals was nearly complete. Serum creatinine, Kim-1 and Ngal were slightly but significantly elevated even in uninjured Sca-1-/- kidneys. Sca-1 constitutively bound both TGFβ receptors I and II in cultured normal proximal tubular epithelial cells. Its genetic loss or silencing lead to constitutive TGFβ receptor-mediated activation of canonical Smad signaling even in the absence of ligand and to KIM-1 expression in the silenced cells. These studies demonstrate that by normally repressing TGFβ-mediated canonical Smad signaling, Sca-1 plays an important in renal epithelial cell homeostasis and in recovery of renal function following ischemic acute kidney injury.

Laboratory testing is the single highest-volume medical activity, making it useful to ask how well one can anticipate whether a given test result will be high, low, or within the reference interval ("normal"). We analyzed 10 years of electronic health records--a total of 69.4 million blood tests--to see how well standard rule-mining techniques can anticipate test results based on patient age and gender, recent diagnoses, and recent laboratory test results. We evaluated rules according to their positive and negative predictive value (PPV and NPV) and area under the receiver-operator characteristic curve (ROC AUCs). Using a stringent cutoff of PPV and/or NPV≥0.95, standard techniques yield few rules for sendout tests but several for in-house tests, mostly for repeat laboratory tests that are part of the complete blood count and basic metabolic panel. Most rules were clinically and pathophysiologically plausible, and several seemed clinically useful for informing pre-test probability of a given result. But overall, rules were unlikely to be able to function as a general substitute for actually ordering a test. Improving laboratory utilization will likely require different input data and/or alternative methods.

Acute kidney injury (AKI) is a common and significant medical problem. Despite the kidney's remarkable regenerative capacity, the mortality rate for the AKI patients is high. Thus, there remains a need to better understand the cellular mechanisms of nephron repair in order to develop new strategies that would enhance the intrinsic ability of kidney tissue to regenerate. Here, using a novel, laser ablation-based, zebrafish model of AKI, we show that collective migration of kidney epithelial cells is a primary early response to acute injury. We also show that cell proliferation is a late response of regenerating kidney epithelia that follows cell migration during kidney repair. We propose a computational model that predicts this temporal relationship and suggests that cell stretch is a mechanical link between migration and proliferation, and present experimental evidence in support of this hypothesis. Overall, this study advances our understanding of kidney repair mechanisms by highlighting a primary role for collective cell migration, laying a foundation for new approaches to treatment of AKI.

Tyrosine recombinases are conserved in the three kingdoms of life. Here we present the first crystal structure of a full-length archaeal tyrosine recombinase, XerA from Pyrococcus abyssi, at 3.0 Å resolution. In the absence of DNA substrate XerA crystallizes as a dimer where each monomer displays a tertiary structure similar to that of DNA-bound Tyr-recombinases. Active sites are assembled in the absence of dif except for the catalytic Tyr, which is extruded and located equidistant from each active site within the dimer. Using XerA active site mutants we demonstrate that XerA follows the classical cis-cleavage reaction, suggesting rearrangements of the C-terminal domain upon DNA binding. Surprisingly, XerA C-terminal αN helices dock in cis in a groove that, in bacterial tyrosine recombinases, accommodates in trans αN helices of neighbour monomers in the Holliday junction intermediates. Deletion of the XerA C-terminal αN helix does not impair cleavage of suicide substrates but prevents recombination catalysis. We propose that the enzymatic cycle of XerA involves the switch of the αN helix from cis to trans packing, leading to (i) repositioning of the catalytic Tyr in the active site in cis and (ii) dimer stabilisation via αN contacts in trans between monomers.

BACKGROUND: Laboratory testing is the single highest-volume medical activity and drives clinical decision-making across medicine. However, the overall landscape of inappropriate testing, which is thought to be dominated by repeat testing, is unclear. Systematic differences in initial vs. repeat testing, measurement criteria, and other factors would suggest new priorities for improving laboratory testing. METHODS: A multi-database systematic review was performed on published studies from 1997-2012 using strict inclusion and exclusion criteria. Over- vs. underutilization, initial vs. repeat testing, low- vs. high-volume testing, subjective vs. objective appropriateness criteria, and restrictive vs. permissive appropriateness criteria, among other factors, were assessed. RESULTS: Overall mean rates of over- and underutilization were 20.6% (95% CI 16.2-24.9%) and 44.8% (95% CI 33.8-55.8%). Overutilization during initial testing (43.9%; 95% CI 35.4-52.5%) was six times higher than during repeat testing (7.4%; 95% CI 2.5-12.3%; P for stratum difference

A large fraction of engineered nanomaterials in consumer and commercial products will reach natural ecosystems. To date, research on the biological impacts of environmental nanomaterial exposures has largely focused on high-concentration exposures in mechanistic lab studies with single strains of model organisms. These results are difficult to extrapolate to ecosystems, where exposures will likely be at low-concentrations and which are inhabited by a diversity of organisms. Here we show adverse responses of plants and microorganisms in a replicated long-term terrestrial mesocosm field experiment following a single low dose of silver nanoparticles (0.14 mg Ag kg(-1) soil) applied via a likely route of exposure, sewage biosolid application. While total aboveground plant biomass did not differ between treatments receiving biosolids, one plant species, Microstegium vimeneum, had 32 % less biomass in the Slurry+AgNP treatment relative to the Slurry only treatment. Microorganisms were also affected by AgNP treatment, which gave a significantly different community composition of bacteria in the Slurry+AgNPs as opposed to the Slurry treatment one day after addition as analyzed by T-RFLP analysis of 16S-rRNA genes. After eight days, N2O flux was 4.5 fold higher in the Slurry+AgNPs treatment than the Slurry treatment. After fifty days, community composition and N2O flux of the Slurry+AgNPs treatment converged with the Slurry. However, the soil microbial extracellular enzymes leucine amino peptidase and phosphatase had 52 and 27% lower activities, respectively, while microbial biomass was 35% lower than the Slurry. We also show that the magnitude of these responses was in all cases as large as or larger than the positive control, AgNO3, added at 4-fold the Ag concentration of the silver nanoparticles.

One-half of the integrin α-subunit Propeller domains contain and extra vWFA domain (αA domain), which mediates integrin binding to extracellular physiologic ligands via its metal-ion-dependent adhesion site (MIDAS). We used electron microscopy to determine the 3D structure of the αA-containing ectodomain of the leukocyte integrin CD11b/CD18 (αMβ2) in its inactive state. A well defined density for αA was observed within a bent ectodomain conformation, while the structure of the ectodomain in complex with the Fab fragment of mAb107, which binds at the MIDAS face of CD11b and stabilizes the inactive state, further revealed that αA is restricted to a relatively small range of orientations relative to the Propeller domain. Using Fab 107 as probe in fluorescent lifetime imaging microscopy (FLIM) revealed that αA is positioned relatively far from the membrane surface in the inactive state, and a systematic orientation search revealed that the MIDAS face would be accessible to extracellular ligand in the inactive state of the full-length cellular integrin. These studies are the first to define the 3D EM structure of an αA-containing integrin ectodomain and to position the ligand-binding face of αA domain in relation to the plasma membrane, providing new insights into current models of integrin activation.

Antibodies' protective, pathological, and therapeutic properties result from their considerable diversity. This diversity is almost limitless in potential, but actual diversity is still poorly understood. Here we use deep sequencing to characterize the diversity of the heavy-chain CDR3 region, the most important contributor to antibody binding specificity, and the constituent V, D, and J segments that comprise it. We find that, during the stepwise D-J and then V-DJ recombination events, the choice of D and J segments exert some bias on each other; however, we find the choice of the V segment is essentially independent of both. V, D, and J segments are utilized with different frequencies, resulting in a highly skewed representation of VDJ combinations in the repertoire. Nevertheless, the pattern of segment usage was almost identical between two different individuals. The pattern of V, D, and J segment usage and recombination was insufficient to explain overlap that was observed between the two individuals' CDR3 repertoires. Finally, we find that while there are a near-infinite number of heavy-chain CDR3s in principle, there are about 3-9 million in the blood of an adult human being.

High-throughput sequencing platforms provide an approach for detecting rare HIV-1 variants and documenting more fully quasispecies diversity. We applied this technology to the V3 loop-coding region of env in samples collected from 4 chronically HIV-infected subjects in whom CCR5 antagonist (vicriviroc [VVC]) therapy failed. Between 25,000-140,000 amplified sequences were obtained per sample. Profound baseline V3 loop sequence heterogeneity existed; predicted CXCR4-using populations were identified in a largely CCR5-using population. The V3 loop forms associated with subsequent virologic failure, either through CXCR4 use or the emergence of high-level VVC resistance, were present as minor variants at 0.8-2.8% of baseline samples. Extreme, rapid shifts in population frequencies toward these forms occurred, and deep sequencing provided a detailed view of the rapid evolutionary impact of VVC selection. Greater V3 diversity was observed post-selection. This previously unreported degree of V3 loop sequence diversity has implications for viral pathogenesis, vaccine design, and the optimal use of HIV-1 CCR5 antagonists.