Divergent RNA viruses infecting sea lice, major ectoparasites of fish.

Sea lice, the major ectoparasites of fish, have significant economic impacts on wild and farmed finfish, and have been implicated in the decline of wild salmon populations. As blood-feeding arthropods, sea lice may also be reservoirs for viruses infecting fish. However, except for two groups of negative-strand RNA viruses within the order Mononegavirales, nothing is known about viruses of sea lice. Here, we used transcriptomic data from three key species of sea lice (Lepeophtheirus salmonis, Caligus clemensi, and Caligus rogercresseyi) to identify 32 previously unknown RNA viruses. The viruses encompassed all the existing phyla of RNA viruses, with many placed in deeply branching lineages that likely represent new families and genera. Importantly, the presence of canonical virus-derived small interfering RNAs (viRNAs) indicates that most of these viruses infect sea lice, even though in some cases their closest classified relatives are only known to infect plants or fungi. We also identified both viRNAs and PIWI-interacting RNAs (piRNAs) from sequences of a bunya-like and two qin-like viruses in C. rogercresseyi. Our analyses showed that most of the viruses found in C. rogercresseyi occurred in multiple life stages, spanning from planktonic to parasitic stages. Phylogenetic analysis revealed that many of the viruses infecting sea lice were closely related to those that infect a wide array of eukaryotes with which arthropods associate, including fungi and parasitic tapeworms, implying that over evolutionary time there has been cross-phylum and cross-kingdom switching of viruses between arthropods and other eukaryotes. Overall, this study greatly expands our view of virus diversity in crustaceans, identifies viruses that infect and replicate in sea lice, and provides evidence that over evolutionary time, viruses have switched between arthropods and eukaryotic hosts in other phyla and kingdoms.

Distinct trophic ecologies of zooplankton size classes are maintained throughout the seasonal cycle.

Marine food webs are strongly size-structured and size-based analysis of communities is a useful approach to evaluate food webs in a way that can be compared across systems. Fatty acid analysis is commonly used to identify diet sources of species, offering a powerful complement to stable isotopes, but is rarely applied to size-structured communities. In this study, we used fatty acids and stable isotopes to characterize size-based variation in prey resources and trophic pathways over a nine-month temperate coastal ocean time series of seven plankton size classes, from > 0.7-µm particulate organic matter through > 2000-µm zooplankton. Zooplankton size classes were generally distinguishable by their dietary fatty acids, while stable isotopes revealed more seasonal variability. Fatty acids of zooplankton were correlated with those of their prey (particulate organic matter and smaller zooplankton) and identified trophic pathways, including widespread ties to the microbial food web. Diatom fatty acids also contributed to zooplankton but fall blooms were more important than spring. Concurrent isotope-based trophic position estimates and fatty acid markers of carnivory showed that some indicators (18:1ω9/18:1ω7) are not consistent across size classes, while others (DHA:EPA) are relatively reliable. Both analysis methods provided distinct information to build a more robust understanding of resource use. For example, fatty acid markers showed that trophic position was likely underestimated in 250-µm zooplankton, probably due to their consumption of protists with low isotopic fractionation factors. Applying fatty acid analysis to a size-structured framework provides more insight into trophic pathways than isotopes alone.

Prostate tumor RON receptor signaling mediates macrophage recruitment to drive androgen deprivation therapy resistance through Gas6-mediated Axl and RON signaling.

BACKGROUND: Androgen deprivation therapy (ADT), or chemical castration, is the first-line therapy for prostate cancer; however, resistance leaves few treatment options. Prostatic tumor-associated macrophages (TAMs) have been shown to promote prostate cancer growth and are abundant in castration-resistant prostate cancer (CRPC), suggesting a role in promoting CRPC. We recently showed a tumor cell-intrinsic mechanism by which RON promotes CRPC. Given previous reports that RON alters prostate cancer cell chemokine production and RON-overexpressing tumors alter macrophage function, we hypothesized that a macrophage-dependent mechanism regulated by tumor cell intrinsic RON also promotes CRPC. METHODS: Using RON-modulated genetically engineered mouse models (GEMMs) and GEMM-derived cell lines and co-cultures with bone marrow-derived macrophages, we show functional and molecular characteristics of signaling pathways in supporting CRPC. Further, we used an unbiased phosphokinase array to identify pathway interactions regulated by RON. Finally, using human prostate cancer cell lines and prostate cancer patient data sets, we show the relevance of our findings to human prostate cancer. RESULTS: Studies herein show that macrophages recruited into the prostate tumor microenvironment (TME) serve as a source for Gas6 secretion which serves to further enhance RON and Axl receptor activation in prostate tumor cells thereby driving CRPC. Further, we show targeting RON and macrophages in a murine model promotes CRPC sensitization to ADT. CONCLUSIONS: We discovered a novel role for the RON receptor in prostate cancer cells in promoting CRPC through the recruitment of macrophages into the prostate TME. Macrophage-targeting agents in combination with RON/Axl inhibition are likely to provide clinical benefits for patients with CRPC.

Experimentally derived trophic enrichment and discrimination factors for Chinook salmon, Oncorhynchus tshawytscha.

RATIONALE: Stable isotope analysis (SIA) can provide important insights into food web structure and is a widely used tool in ecological conservation and management. It has recently been augmented by compound-specific stable isotope analysis of amino acids (CSIA-AA), an innovation that can provide greater precision when analyzing trophic level and food web connectivity. The utility of SIA rests on confidence in its constituent parameters such as the trophic enrichment factor (TEF). There is increasing emphasis on the need to experimentally derive species and tissue specific TEFs for studies utilizing SIA. Chinook salmon, Oncorhynchus tshawytscha, is a species with high potential for study using SIA due to the difficulty in observing its ecology during its marine phase and the significance of the conservation consequences of recent population declines. METHODS: Bulk and amino acid-specific TEFs were determined for juvenile and adult Chinook salmon fed specific diets. Three controlled feeding studies were performed: adult salmon were fed a biofeed, juvenile salmon were fed a biofeed, and juvenile salmon were fed krill. Bulk and compound-specific stable isotope data were collected from diet samples and from salmon muscle tissue after a minimum of 8 weeks of controlled feeding. Bulk isotope signatures were measured using EA-IRMS and CSIA-AA signatures using GC/C-IRMS, allowing the TEFs to be calculated. RESULTS: The bulk isotope TEFs were higher than those predicted for similar marine organisms and averaged 3.5‰ for ∆15 N and 1.3‰ for ∆13 C. The TEFs derived for nitrogen isotopes of amino acids were in line with expectations for this approach: the mean value for ∆15 NGlu - ∆15 NPhe was 7.06‰ and, using a multi-AA approach, the value for ∆15 NTrophic - ∆15 NSource was 6.67‰. For carbon isotopes of amino acids, the derived TEFs of Iso, Leu and Phe were near 0‰, as was that of Met, supporting their use of as source amino acids in future CSIA studies. CONCLUSIONS: This study presents Chinook salmon-specific TEFs for bulk and amino acid SIA. It supports the application of future research applying SIA to the study of Chinook salmon and validates previous research on species-specific TEFs.

Using data assimilation to train a hybrid forecast system that combines machine-learning and knowledge-based components.

We consider the problem of data-assisted forecasting of chaotic dynamical systems when the available data are in the form of noisy partial measurements of the past and present state of the dynamical system. Recently, there have been several promising data-driven approaches to forecasting of chaotic dynamical systems using machine learning. Particularly promising among these are hybrid approaches that combine machine learning with a knowledge-based model, where a machine-learning technique is used to correct the imperfections in the knowledge-based model. Such imperfections may be due to incomplete understanding and/or limited resolution of the physical processes in the underlying dynamical system, e.g., the atmosphere or the ocean. Previously proposed data-driven forecasting approaches tend to require, for training, measurements of all the variables that are intended to be forecast. We describe a way to relax this assumption by combining data assimilation with machine learning. We demonstrate this technique using the Ensemble Transform Kalman Filter to assimilate synthetic data for the three-variable Lorenz 1963 system and for the Kuramoto-Sivashinsky system, simulating a model error in each case by a misspecified parameter value. We show that by using partial measurements of the state of the dynamical system, we can train a machine-learning model to improve predictions made by an imperfect knowledge-based model.

MST1R (RON) expression is a novel prognostic biomarker for metastatic progression in breast cancer patients.

PURPOSE: This study evaluates the prognostic significance of MST1R (RON) expression in breast cancer with respect to disease progression, long-term survival, subtype, and association with conventional prognostic factors. METHODS: The approach includes interrogation of survival and tumor staging with paired MST1R RNA expression from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) datasets. Protein expression evaluation was performed using immunohistochemistry (IHC) staining of MST1R on breast cancer tissue samples from the Cancer Diagnosis Program Breast Cancer Progression tissue microarray and locally obtained breast tumor tissue samples analyzed with paired survival, metastasis, and subtype. RESULTS: Data from TCGA (n = 774) show poorer relapse-free survival (RFS) in patients with high MST1R expression (P = 0.32) and no difference in MST1R expression based on tumor stage (P = 0.77) or nodal status (P = 0.94). Patients in the GEO-derived Kaplan-Meier Plotter microarray dataset demonstrate the association of MST1R and poorer overall survival (n = 1402, P = 0.018) and RFS in patients receiving chemotherapy (n = 798, P = 0.041). Patients with high MST1R expression display worse overall survival (P = 0.01) and receiver operator characteristic (ROC) analysis demonstrate the predictive capacity of increased MST1R with early death (P = 0.0017) in IHC-stained samples. Paired IHC-stained breast tumor samples from the primary versus metastatic site show MST1R expression is associated with metastatic progression (P = 0.032), and ROC analysis supports the predictive capacity of MST1R in metastatic progression (P = 0.031). No associations of MST1R with estrogen receptor (ER), progesterone receptor (PR), both ER and PR, HER2 positivity, or triple-negativity were found (P = 0.386, P = 0.766, P = 0.746, P = 0.457, P = 0.947, respectively). CONCLUSIONS: MST1R expression has prognostic value in breast cancer with respect to survival and metastatic progression. MST1R expression is not associated with tumor stage, nodal status, or subtype.

Separation of chaotic signals by reservoir computing.

We demonstrate the utility of machine learning in the separation of superimposed chaotic signals using a technique called reservoir computing. We assume no knowledge of the dynamical equations that produce the signals and require only training data consisting of finite-time samples of the component signals. We test our method on signals that are formed as linear combinations of signals from two Lorenz systems with different parameters. Comparing our nonlinear method with the optimal linear solution to the separation problem, the Wiener filter, we find that our method significantly outperforms the Wiener filter in all the scenarios we study. Furthermore, this difference is particularly striking when the component signals have similar frequency spectra. Indeed, our method works well when the component frequency spectra are indistinguishable-a case where a Wiener filter performs essentially no separation.

Combining machine learning with knowledge-based modeling for scalable forecasting and subgrid-scale closure of large, complex, spatiotemporal systems.

We consider the commonly encountered situation (e.g., in weather forecast) where the goal is to predict the time evolution of a large, spatiotemporally chaotic dynamical system when we have access to both time series data of previous system states and an imperfect model of the full system dynamics. Specifically, we attempt to utilize machine learning as the essential tool for integrating the use of past data into predictions. In order to facilitate scalability to the common scenario of interest where the spatiotemporally chaotic system is very large and complex, we propose combining two approaches: (i) a parallel machine learning prediction scheme and (ii) a hybrid technique for a composite prediction system composed of a knowledge-based component and a machine learning-based component. We demonstrate that not only can this method combining (i) and (ii) be scaled to give excellent performance for very large systems but also that the length of time series data needed to train our multiple, parallel machine learning components is dramatically less than that necessary without parallelization. Furthermore, considering cases where computational realization of the knowledge-based component does not resolve subgrid-scale processes, our scheme is able to use training data to incorporate the effect of the unresolved short-scale dynamics upon the resolved longer-scale dynamics (subgrid-scale closure).

Actinobacillus vicugnae sp. nov., isolated from alpaca (Vicugna pacos).

Ten strains of an Actinobacillus-like organism were isolated from alpaca (Vicugna pacos) in the UK over a period of 5 years, with no known epidemiological linkages. The isolates are distinct, based on both phenotype and genotype, from any previously described Actinobacillus species. Molecular analysis, based on 16S rRNA, rpoB and infB gene sequences, placed the isolates as a novel, early branching, lineage within the currently recognised Actinobacillus sensu stricto. In agreement with the results of the single-gene analysis, average nucleotide identity values, based on whole genome sequences, showed very similar identities to a number of members of the Actinobacillus sensu stricto notably Actinobacillus equuli, Actinobacillus suis and Actinobacillus ureae. At least two phenotypic characteristics differentiate the alpaca isolates from other Actinobacillus sensu stricto species, and from taxa likely falling within this group but awaiting formal species description, with Actinobacillus anseriformium and A. equulisubsp. haemolyticus being the most closely related phenotypically. The alpaca isolates can be differentiated from A. anseriformium by production of ß-galactosidase (ONPG) and acid from raffinose, and from A. equulisubsp. haemolyticus by production of acid from d-sorbitol and failure to produce acid from d-xylose. Isolates were obtained from multiple sites in alpaca including respiratory tract, alimentary tract and internal organs although further evidence is required to understand any pathogenic significance. Based on the results of characterization described here, it is proposed that the isolates constitute a novel species, Actinobacillus vicugnae sp. nov. The type strain is W1618T (LMG30745T NCTC14090T) isolated in the UK in 2012 from oesophageal ulceration in an alpaca (Vicugna pacos).

Dynamics of analog logic-gate networks for machine learning.

We describe the continuous-time dynamics of networks implemented on Field Programable Gate Arrays (FPGAs). The networks can perform Boolean operations when the FPGA is in the clocked (digital) mode; however, we run the programed FPGA in the unclocked (analog) mode. Our motivation is to use these FPGA networks as ultrafast machine-learning processors, using the technique of reservoir computing. We study both the undriven dynamics and the input response of these networks as we vary network design parameters, and we relate the dynamics to accuracy on two machine-learning tasks.

Anthropogenic disturbance homogenizes seagrass fish communities.

Anthropogenic activities have led to the biotic homogenization of many ecological communities, yet in coastal systems this phenomenon remains understudied. In particular, activities that locally affect marine habitat-forming foundation species may perturb habitat and promote species with generalist, opportunistic traits, in turn affecting spatial patterns of biodiversity. Here, we quantified fish diversity in seagrass communities across 89 sites spanning 6° latitude along the Pacific coast of Canada, to test the hypothesis that anthropogenic disturbances homogenize (i.e., lower beta-diversity) assemblages within coastal ecosystems. We test for patterns of biotic homogenization at sites within different anthropogenic disturbance categories (low, medium, and high) at two spatial scales (within and across regions) using both abundance- and incidence-based beta-diversity metrics. Our models provide clear evidence that fish communities in high anthropogenic disturbance seagrass areas are homogenized relative to those in low disturbance areas. These results were consistent across within-region comparisons using abundance- and incidence-based measures of beta-diversity, and in across-region comparisons using incidence-based measures. Physical and biotic characteristics of seagrass meadows also influenced fish beta-diversity. Biotic habitat characteristics including seagrass biomass and shoot density were more differentiated among high disturbance sites, potentially indicative of a perturbed environment. Indicator species and trait analyses revealed fishes associated with low disturbance sites had characteristics including stenotopy, lower swimming ability, and egg guarding behavior. Our study is the first to show biotic homogenization of fishes across seagrass meadows within areas of relatively high human impact. These results support the importance of targeting conservation efforts in low anthropogenic disturbance areas across land- and seascapes, as well as managing anthropogenic impacts in high activity areas.

Model-Free Prediction of Large Spatiotemporally Chaotic Systems from Data: A Reservoir Computing Approach.

We demonstrate the effectiveness of using machine learning for model-free prediction of spatiotemporally chaotic systems of arbitrarily large spatial extent and attractor dimension purely from observations of the system's past evolution. We present a parallel scheme with an example implementation based on the reservoir computing paradigm and demonstrate the scalability of our scheme using the Kuramoto-Sivashinsky equation as an example of a spatiotemporally chaotic system.

Attractor reconstruction by machine learning.

A machine-learning approach called "reservoir computing" has been used successfully for short-term prediction and attractor reconstruction of chaotic dynamical systems from time series data. We present a theoretical framework that describes conditions under which reservoir computing can create an empirical model capable of skillful short-term forecasts and accurate long-term ergodic behavior. We illustrate this theory through numerical experiments. We also argue that the theory applies to certain other machine learning methods for time series prediction.

Hybrid forecasting of chaotic processes: Using machine learning in conjunction with a knowledge-based model.

A model-based approach to forecasting chaotic dynamical systems utilizes knowledge of the mechanistic processes governing the dynamics to build an approximate mathematical model of the system. In contrast, machine learning techniques have demonstrated promising results for forecasting chaotic systems purely from past time series measurements of system state variables (training data), without prior knowledge of the system dynamics. The motivation for this paper is the potential of machine learning for filling in the gaps in our underlying mechanistic knowledge that cause widely-used knowledge-based models to be inaccurate. Thus, we here propose a general method that leverages the advantages of these two approaches by combining a knowledge-based model and a machine learning technique to build a hybrid forecasting scheme. Potential applications for such an approach are numerous (e.g., improving weather forecasting). We demonstrate and test the utility of this approach using a particular illustrative version of a machine learning known as reservoir computing, and we apply the resulting hybrid forecaster to a low-dimensional chaotic system, as well as to a high-dimensional spatiotemporal chaotic system. These tests yield extremely promising results in that our hybrid technique is able to accurately predict for a much longer period of time than either its machine-learning component or its model-based component alone.

Maternal diethylhexyl phthalate exposure affects adiposity and insulin tolerance in offspring in a PCNA-dependent manner.

The ubiquitous plasticizer, diethylhexyl phthalate (DEHP), is a known endocrine disruptor. However, DEHP exposure effects are not well understood. Changes in industrial and agricultural practices have resulted in increased prevalence of DEHP exposure and has coincided with the heightened occurrence of metabolic syndrome and obesity. DEHP and its metabolites are detected in the umbilical cord blood of newborns; however, the prenatal and perinatal effects of DEHP exposure have not been intensively studied. Previously, we discovered that phosphorylation (p) of proliferating cell nuclear antigen (PCNA) at tyrosine 114 (Y114) is required for adipogenesis and diet-induced obesity in mice. Here, we show the unique ability of DEHP to induce p-Y114 in PCNA in vitro. We also show that while DEHP promotes adipogenesis of wild type (WT) murine embryonic fibroblasts, mutation of Y114 to phenylalanine (Y114F) in PCNA blocked adipocyte differentiation. Given the induction of p-Y114 in PCNA by DEHP and the relationship to obesity, WT and Y114F PCNA mice were exposed to DEHP during gestation or lactation, followed by high fat diet feeding. Paradoxically, in utero exposure of Y114F PCNA females to DEHP led to a significant increase in body mass and was associated with augmented expression of PPARγ, a critical regulator of obesity, compared to WT controls. In utero exposure of WT mice to DEHP led to insulin sensitivity while Y114F mutation ablated this phenotype, indicating that PCNA is an important regulator of early DEHP exposure and ensuing metabolic phenotypes.

Using machine learning to replicate chaotic attractors and calculate Lyapunov exponents from data.

We use recent advances in the machine learning area known as "reservoir computing" to formulate a method for model-free estimation from data of the Lyapunov exponents of a chaotic process. The technique uses a limited time series of measurements as input to a high-dimensional dynamical system called a "reservoir." After the reservoir's response to the data is recorded, linear regression is used to learn a large set of parameters, called the "output weights." The learned output weights are then used to form a modified autonomous reservoir designed to be capable of producing an arbitrarily long time series whose ergodic properties approximate those of the input signal. When successful, we say that the autonomous reservoir reproduces the attractor's "climate." Since the reservoir equations and output weights are known, we can compute the derivatives needed to determine the Lyapunov exponents of the autonomous reservoir, which we then use as estimates of the Lyapunov exponents for the original input generating system. We illustrate the effectiveness of our technique with two examples, the Lorenz system and the Kuramoto-Sivashinsky (KS) equation. In the case of the KS equation, we note that the high dimensional nature of the system and the large number of Lyapunov exponents yield a challenging test of our method, which we find the method successfully passes.

Reservoir observers: Model-free inference of unmeasured variables in chaotic systems.

Deducing the state of a dynamical system as a function of time from a limited number of concurrent system state measurements is an important problem of great practical utility. A scheme that accomplishes this is called an "observer." We consider the case in which a model of the system is unavailable or insufficiently accurate, but "training" time series data of the desired state variables are available for a short period of time, and a limited number of other system variables are continually measured. We propose a solution to this problem using networks of neuron-like units known as "reservoir computers." The measurements that are continually available are input to the network, which is trained with the limited-time data to output estimates of the desired state variables. We demonstrate our method, which we call a "reservoir observer," using the Rössler system, the Lorenz system, and the spatiotemporally chaotic Kuramoto-Sivashinsky equation. Subject to the condition of observability (i.e., whether it is in principle possible, by any means, to infer the desired unmeasured variables from the measured variables), we show that the reservoir observer can be a very effective and versatile tool for robustly reconstructing unmeasured dynamical system variables.

The Effect of Intermittent Head Cooling on Aerobic Performance in the Heat.

Thermoregulation is critical for athletes, particularly those for those who must perform in the heat. Most strategies aimed at reducing heat stress have cooled participants before or during activity. The objective of this study is to investigate whether seven minutes of head cooling applied between bouts of aerobic exercise in hot (35 ± 1.0 °C) and dry (14.68 ±4.29% rh) environmental conditions could positively effect participants peak power output (PP) on a maximal effort graded exercise test (GXT). Twenty-two recreational active men ages 18 to 23 (19.8 ± 1.6 yrs.) completed three performance trials over a 21 day period. During the first trial, participants were familiarized with procedures and completed a maximal effort GXT on a cycle ergometer to establish maximal baseline performances. The second and third trials, which were counterbalanced, consisted of a cooling and placebo condition. During both of these trials, participants cycled 40 minutes at 65% of their maximum VO2, in hot (35 ± 1.0 °C) and dry (17-20% rh) environmental conditions. Immediately after this initial bout of activity, participants were given seven minutes of recovery in which head cooling was applied during the cooling condition and withheld during the placebo condition. Participants then completed a maximal effort GXT. Significant differences (p < 0.001) in participants peak power output (W) were measured when cooling was applied compared to the placebo condition (304.23(W) ± 26.19(W) cooling, 291.68(W) ± 26.04(W) placebo). These results suggest that a relatively brief period of intermittent cooling may enhance subsequent aerobic performance.

Defining chaos.

In this paper, we propose, discuss, and illustrate a computationally feasible definition of chaos which can be applied very generally to situations that are commonly encountered, including attractors, repellers, and non-periodically forced systems. This definition is based on an entropy-like quantity, which we call "expansion entropy," and we define chaos as occurring when this quantity is positive. We relate and compare expansion entropy to the well-known concept of topological entropy to which it is equivalent under appropriate conditions. We also present example illustrations, discuss computational implementations, and point out issues arising from attempts at giving definitions of chaos that are not entropy-based.

Stabilizing machine learning prediction of dynamics: Novel noise-inspired regularization tested with reservoir computing.

Recent work has shown that machine learning (ML) models can skillfully forecast the dynamics of unknown chaotic systems. Short-term predictions of the state evolution and long-term predictions of the statistical patterns of the dynamics ("climate") can be produced by employing a feedback loop, whereby the model is trained to predict forward only one time step, then the model output is used as input for multiple time steps. In the absence of mitigating techniques, however, this feedback can result in artificially rapid error growth ("instability"). One established mitigating technique is to add noise to the ML model training input. Based on this technique, we formulate a new penalty term in the loss function for ML models with memory of past inputs that deterministically approximates the effect of many small, independent noise realizations added to the model input during training. We refer to this penalty and the resulting regularization as Linearized Multi-Noise Training (LMNT). We systematically examine the effect of LMNT, input noise, and other established regularization techniques in a case study using reservoir computing, a machine learning method using recurrent neural networks, to predict the spatiotemporal chaotic Kuramoto-Sivashinsky equation. We find that reservoir computers trained with noise or with LMNT produce climate predictions that appear to be indefinitely stable and have a climate very similar to the true system, while the short-term forecasts are substantially more accurate than those trained with other regularization techniques. Finally, we show the deterministic aspect of our LMNT regularization facilitates fast reservoir computer regularization hyperparameter tuning.