Increase In Prediction Error Research Articles

Integration of Sense and Control for Uncertain Systems Based on Delayed Feedback Active Inference.

Asa result of the time lag in transmission, the data obtained by the sensor is delayed and does not reflect the state at the current moment. The effects of input delay are often overlooked in active inference (AIF), which may lead to significant deviations in state estimation and increased prediction errors, particularly when the system is subjected to a sudden external stimulus. In this paper, a theoretical framework of delayed feedback active inference (DAIF) is proposed to enhance the applicability of AIF to real systems. The probability model of DAIF is defined by incorporating a control distribution into that of AIF. The free energy of DAIF is defined as the sum of the quadratic state, sense, and control prediction error. A predicted state derived from previous states is defined and introduced as the expectation of the prior distribution of the real-time state. A proportional-integral (PI)-like control based on the predicted state is taken to be the expectation of DAIF preference control, whose gain coefficient is inversely proportional to the measurement accuracy variance. To adaptively compensate for external disturbances, a second-order inverse variance accuracy replaces the fixed sensory accuracy of preference control. The simulation results of the trajectory tracking control of a quadrotor unmanned aerial vehicle (UAV) show that DAIF performs better than AIF in state estimation and disturbance resistance.

Calibration modeling of drilling fluid rheological parameters in variable temperature environment based on support vector machine.

Traditional measurement of drilling fluid rheological parameters suffers from significant lag due to the inability of the instruments to promptly capture real-time parameters of the drilling fluid. These measurement models are typically constructed based on fixed temperature conditions and empirical formulas, rendering them inadequate for complex temperature gradient environments. Consequently, this limitation results in increased prediction errors, severely compromising the precise monitoring of drilling fluid performance. Aiming at the problems of low accuracy and poor stability of drilling fluid measurements under variable temperature conditions, a support vector machine-based calibration model for drilling fluid rheological parameters in a variable temperature environment is proposed in this paper. First, the measurement principle of the double-tube differential pressure rheology real-time measurement device is analyzed. The relationship between shear stress and shear rate is then established using differential pressure sensor and flow rate data. Utilizing the gray wolf optimization algorithm to optimize the kernel function weights and parameters, an SVM-based calibration model for predicting drilling fluid rheology correction parameters is constructed. Finally, a real-time monitoring platform for drilling fluid is developed. Experimental results show that the maximum relative errors for the predictions of apparent viscosity, plastic viscosity, and yield point are within ±5%, with coefficients of determination (R2) all greater than 0.95. These results validate the effectiveness of the proposed method in accurately monitoring the rheological performance of drilling fluids.

EXPLORING CONCUSSION RECOVERY THROUGH THE LENS OF THE FREE ENERGY PRINCIPLE AND MARKOV BLANKET THEORY

This paper delves into the application of the Free Energy Principle (FEP) and the concept of the Markov blanket in understanding the neurocognitive implications of concussion. The FEP, a unifying theory in neuroscience, posits that the brain functions to minimize free energy, equating to a reduction in surprise or uncertainty regarding sensory inputs. The Markov blanket, defining the boundary between a system and its environment, is integral to this framework, particularly in understanding how the brain processes and responds to sensory information. We explore how a concussion might disrupt the brain's predictive processing and its ability to minimize free energy, leading to increased prediction errors and cognitive deficits. This disruption is hypothesized to manifest as an inability to accurately predict sensory inputs, resulting in impaired cognitive functions post-concussion. The paper also discusses the potential of neuroplasticity in concussion recovery, framed within the FEP as the brain's effort to re-establish minimized free energy under new constraints. We propose a methodological approach combining neuroimaging and computational modeling to empirically test these hypotheses. This theoretical exploration offers novel insights into the mechanisms underlying concussion-induced cognitive impairments and suggests new avenues for therapeutic intervention and rehabilitation strategies.

Transition role of entangled data in quantum machine learning

Entanglement serves as the resource to empower quantum computing. Recent progress has highlighted its positive impact on learning quantum dynamics, wherein the integration of entanglement into quantum operations or measurements of quantum machine learning (QML) models leads to substantial reductions in training data size, surpassing a specified prediction error threshold. However, an analytical understanding of how the entanglement degree in data affects model performance remains elusive. In this study, we address this knowledge gap by establishing a quantum no-free-lunch (NFL) theorem for learning quantum dynamics using entangled data. Contrary to previous findings, we prove that the impact of entangled data on prediction error exhibits a dual effect, depending on the number of permitted measurements. With a sufficient number of measurements, increasing the entanglement of training data consistently reduces the prediction error or decreases the required size of the training data to achieve the same prediction error. Conversely, when few measurements are allowed, employing highly entangled data could lead to an increased prediction error. The achieved results provide critical guidance for designing advanced QML protocols, especially for those tailored for execution on early-stage quantum computers with limited access to quantum resources.

Risk-Taking Is Associated with Decreased Subjective Value Signals and Increased Prediction Error Signals in the Hot Columbia Card Task.

It remains a pressing concern to understand how neural computations relate to risky decisions. However, most observations of brain-behavior relationships in the risk-taking domain lack a rigorous computational basis or fail to emulate of the dynamic, sequential nature of real-life risky decision-making. Recent advances emphasize the role of neural prediction error (PE) signals. We modeled, according to prospect theory, the choices of n = 43 human participants (33 females, 10 males) performing an EEG version of the hot Columbia Card Task, featuring rounds of sequential decisions between stopping (safe option) and continuing with increasing odds of a high loss (risky option). Single-trial regression EEG analyses yielded a subjective value signal at centroparietal (300-700 ms) and frontocentral (>800 ms) electrodes and in the delta band, as well as PE signals tied to the feedback-related negativity, P3a, and P3b, and in the theta band. Higher risk preference (total number of risky choices) was linked to attenuated subjective value signals but increased PE signals. Higher P3-like activity associated with the most positive PE in each round predicted stopping in the present round but not risk-taking in the subsequent round. Our findings indicate that decreased representation of decision values and increased sensitivity to winning despite low odds (positive PE) facilitate risky choices at the subject level. Strong neural responses when gains are least expected (the most positive PE on each round) adaptively contribute to safer choices at the trial-by-trial level but do not affect risky choice at the round-by-round level.

Development and comparison of forensic interval age prediction models by statistical and machine learning methods based on the methylation rates of ELOVL2 in blood DNA

Age estimation can be useful information for narrowing down candidates of unidentified donors in criminal investigations. Various age estimation models based on DNA methylation biomarkers have been developed for forensic usage in the past decade. However, many of these models using ordinary least squares regression cannot generate an appropriate estimation due to the deterioration in prediction accuracy caused by an increased prediction error in older age groups. In the present study, to address this problem, we developed age estimation models that set an appropriate prediction interval for all age groups by two approaches: a statistical method using quantile regression (QR) and a machine learning method using an artificial neural network (ANN). Methylation datasets (n = 1280, age 0–91 years) of the promoter for the gene encoding ELOVL fatty acid elongase 2 were used to develop the QR and ANN models. By validation using several test datasets, both models were shown to enlarge prediction intervals in accordance with aging and have a high level of correct prediction (>90 %) for older age groups. The QR and ANN models also generated a point age prediction with high accuracy. The ANN model enabled a prediction with a mean absolute error (MAE) of 5.3 years and root mean square error (RMSE) of 7.3 years for the test dataset (n = 549), which were comparable to those of the QR model (MAE = 5.6 years, RMSE = 7.8 years). Their applicability to casework was also confirmed using bloodstain samples stored for various periods of time (1–14 years), indicating the stability of the models for aged bloodstain samples. From these results, it was considered that the proposed models can provide more useful and effective age estimation in forensic settings.

Feasibility and performance of cross-clone Raman calibration models in CHO cultivation.

Raman spectroscopy is widely used in monitoring and controlling cell cultivations for biopharmaceutical drug manufacturing. However, its implementation for culture monitoring in the cell line development stage has received little attention. Therefore, the impact of clonal differences, such as productivity and growth, on the prediction accuracy and transferability of Raman calibration models is not yet well described. Raman OPLS models were developed for predicting titer, glucose and lactate using eleven CHO clones from a single cell line. These clones exhibited diverse productivity and growth rates. The calibration models were evaluated for clone-related biases using clone-wise linear regression analysis on cross validated predictions. The results revealed that clonal differences did not affect the prediction of glucose and lactate, but titer models showed a significant clone-related bias, which remained even after applying variable selection methods. The bias was associated with clonal productivity and lead to increased prediction errors when titer models were transferred to cultivations with productivity levels outside the range of their training data. The findings demonstrate the feasibility of Raman-based monitoring of glucose and lactate in cell line development with high accuracy. However, accurate titer prediction requires careful consideration of clonal characteristics during model development.

Force Field Analysis Software and Tools (FFAST): Assessing Machine Learning Force Fields under the Microscope.

As the sophistication of machine learning force fields (MLFF) increases to match the complexity of extended molecules and materials, so does the need for tools to properly analyze and assess the practical performance of MLFFs. To go beyond average error metrics and into a complete picture of a model's applicability and limitations, we developed FFAST (force field analysis software and tools): a cross-platform software package designed to gain detailed insights into a model's performance and limitations, complete with an easy-to-use graphical user interface. The software allows the user to gauge the performance of any molecular force field,─such as popular state-of-the-art MLFF models, ─ on various popular data set types, providing general prediction error overviews, outlier detection mechanisms, atom-projected errors, and more. It has a 3D visualizer to find and picture problematic configurations, atoms, or clusters in a large data set. In this paper, the example of the MACE and NequIP models is used on two data sets of interest [stachyose and docosahexaenoic acid (DHA)]─to illustrate the use cases of the software. With this, it was found that carbons and oxygens involved in or near glycosidic bonds inside the stachyose molecule present increased prediction errors. In addition, prediction errors on DHA rise as the molecule folds, especially for the carboxylic group at the edge of the molecule. We emphasize the need for a systematic assessment of MLFF models for ensuring their successful application to the study of dynamics of molecules and materials.

Ability of three dairy feed evaluation systems to predict postruminal outflows of nitrogenous compounds in dairy cows: A meta-analysis

Adequate prediction of postruminal outflow of protein fractions is the starting point for the determination of metabolizable protein supply in dairy cows. The objective of this meta-analysis was to compare the performance of 3 dairy feed evaluation systems (National Research Council [NRC], Cornell Net Protein and Carbohydrate System [CNCPS], and National Academies of Sciences, Engineering and Medicine [NASEM]) to predict outflows (g/d) of nonammonia nitrogren (NAN), microbial N (MiN), and nonammonia nonmicrobial N (NANMN). Predictions of rumen degradabilities (% of nutrient) of protein (RDP), NDF, and starch were also evaluated. The data set included 1,294 treatment means from 312 digesta flow studies. The 3 feed evaluation systems were compared using the concordance correlation coefficient (CCC), the ratio of root mean square prediction error (RMSPE) on standard deviation of observed values (RSR), and the slope between observed and predicted values. Mean and linear biases were deemed biologically relevant and are discussed if higher than a threshold of 5% of the mean of observed values. The comparisons were done on observed values adjusted or not for the study effect; the adjustment had a small effect on the mean bias but the linear bias reflected a response to a dietary change rather than absolute predictions. For the absolute predictions of NAN and MiN, CNCPS had the best-fit statistics (8% greater CCC; 6% lower RMSPE) without any bias; NRC and NASEM underpredicted NAN and MiN, and NASEM had an additional linear bias indicating that the underprediction of MiN increased at increased predictions. For NANMN, fit statistics were similar among the 3 feed evaluation systems with no mean bias; however, the linear bias with NRC and CNCPS indicated underprediction at low predictions and overprediction at elevated predictions. On average, the CCC were smaller and RSR ratios were greater for MiN versus NAN indicating increased prediction errors for MiN. For NAN responses to a dietary change, CNCPS also had the best predictions, although the mean bias with NASEM was not biologically relevant and the 3 feed evaluation systems did not present a linear bias. However, CNCPS, but not the 2 other feed evaluation systems, presented a linear bias for MiN, with responses being overpredicted at increased predictions. For NANMN, responses were overpredicted at increased predictions for the 3 feed evaluation systems, but to a lesser extent with NASEM. The site of sampling had an effect on the mean bias of MiN and NANMN in the 3 feed evaluation systems. The mean bias of MiN was higher in omasal than duodenal studies in the 3 feed evaluation systems (from 55 to 61 g/d) and this mean bias was twice as large when 15N labeling was used as a microbial marker compared with purines. Such a difference was not observed for duodenal studies. The reasons underlying these systematic differences are not clear as the type of measurements used in the current meta-analysis does not allow to delineate if one site or one microbial marker is yielding the "true" postruminal N outflows. Rumen degradabilities of protein was underpredicted with CNCPS, and RDP responses to a dietary change was underpredicted by the 3 feed evaluation systems with increased RDP predictions. Rumen degradability of NDF was underpredicted and had poor fit statistics for NASEM compared with CNCPS. Fit statistics were similar between CNCPS and NASEM for rumen degradability of starch, but with an underprediction of the response with NASEM and absolute values being overpredicted with CNCPS. Multivariate regression analyses showed that diet characteristics were correlated with prediction errors of N outflows in each feed evaluation system. Globally, compared with NAN and NANMN, residuals of MiN were correlated with several moderators in the 3 feed evaluation systems reflecting the complexity to measure and model this outflow. In addition, residuals of NANMN were correlated positively with RDP suggesting an overestimation of this parameter. In conclusion, although progress is still to be made to improve equations predicting postruminal N outflows, the current feed evaluation systems provide sufficient precision and accuracy to predict postruminal outflows of N fractions.

Bayesian model selection considering model complexity using stochastic filtering

AbstractThis paper proposes a Bayesian model selection framework to determine optimal, parsimonious models for given building structures under ground motion excitation using stochastic filtering. Structural system identification at a regional scale after an earthquake is nominally a monumental task due to its inherently high computational cost. This challenge is currently addressed in the literature by using simplified structural models that bear analytical solutions, such as Timoshenko or shear beams, and shear building models. However, a low computational effort usually leads to an increased prediction error and consequently, a higher model uncertainty. This poses the dilemma of prediction accuracy versus model complexity when selecting a model class to represent a building structure. The proposed framework selects the model class that strikes the best balance between the two. To this end, the notion of cumulative evidence is introduced here as the integral of local evidence over the ground motion duration, which is then formulated as the difference between the cumulative likelihood of the observed measurements and the cumulative penalty. The likelihood measure promotes models whose predictions better match the observations. This is counteracted by the penalty, which is devised as an “Ockham factor,” and penalizes models with higher information gains due to their higher complexities, for example, a larger number of parameters. The proposed approach also yields the best tuning parameters of the stochastic filter for each model class and the best initial values of the identification parameters. The proposed framework is verified using a synthetic example, and validated using recorded data from the Millikan Library building in California and the ANX building in Japan.

Revealing spatial variability of groundwater level in typical ecosystems of the Tarim Basin through ensemble algorithms and limited observations

Owing to the lack of hydrological monitoring stations and groundwater observations, particularly for natural habitats, the distribution of groundwater along the Mountainous (alluvial fan)-Oasis-Desert System (MODS) in the middle reaches of the Tarim River has not yet been systematically investigated. The performance of ensemble algorithms with multi-source/multi-scale (30 m, 90 m, 250 m, 500 m, and 1000 m) remote sensing data to predict the groundwater levels of the MODS of the Tarim Basin should be investigated. To improve the current knowledge on the spatial distribution of groundwater in this region, we conducted a study on the applicability of ensemble learning algorithms for groundwater assessments in typical regions of dryland, namely the Oasis-Taklimakan Desert (OTD), Nature land in Oasis system (NLOS), Irrigation area (IA), and Oasis system (OS). The results showed that the R2 values of the groundwater level prediction accuracy for the four ecosystems were 0.92, 0.96, 0.86, and 0.50, respectively, and their corresponding optimal ensemble algorithms were AR_GRBFN, AR_RF, RotationForest_MLP, and AR_ GRBFN. The scales corresponding to the optimal prediction accuracy of OTD, NLOS, IA, and OS were 250 m, 30 m, 250 m, and 90 m, respectively, and the effect of the scale on their respective groundwater level prediction accuracies (maximum value compared to minimum value) were 11.48%, 21.63%, 26.73%, and 10.59%. Random Subspace, Rotation Forest, and Additive Regression improved the prediction accuracy of the base learning method by 76%, 72%, and 64%, respectively, followed by Bagging, whereas Dagging greatly increased prediction errors. Rotation Forest and Random Subspace showed stable performances and guaranteed relatively low prediction errors. Among all ensemble algorithms, Additive Regression helped the base learners obtain relatively optimal prediction accuracies with high probabilities. Except for the OS system, groundwater level could be predicted with much greater accuracy at depths of >8 m than at <8 m. The contribution analysis showed that topography and land-use patterns controlled the spatial distribution of groundwater across MODS. The ensemble learning algorithm showed good performance using multi-source and multi-scale data.

Real-time prediction of stomach motions based upon gastric contraction and breathing models

Objective. Precision radiation therapy requires managing motions of organs at risk that occur during treatment. While methods have been developed for real-time respiratory motion tracking, non-breathing intra-fractional variations (including gastric contractile motion) have seen little attention to date. The purpose of this study is to develop a cyclic gastric contractile motion prediction model to support real-time management during radiotherapy. Approach. The observed short-term reproducibility of gastric contractile motion permitted development of a prediction model that (1) extracts gastric contraction motion phases from few minutes of golden angle stack of stars scanning (at patient positioning), (2) estimate gastric phase of real-time sampled data acquired during treatment delivery to these reconstructed phases and (3) predicting future gastric phase by linear extrapolation using estimation results from step 2 to account for processing and system latency times. Model was evaluated on three parameters including training time window for step 1, number of spokes for real-time sampling data in step 2 and future prediction time. Main results. The model was tested on a population of 20 min data samples from 25 scans from 15 patients. The mean prediction error with 10 spokes and 2 min training was 0.3 ± 0.1 mm (0.1–0.7 mm) with 5.1 s future time, slowly rising to 0.6 ± 0.2 mm (0.2–1.1 mm) for 6.8 s future time and then increasing rapidly for longer forward predictions, for an average 3.6 ± 0.5 mm (2.8–4.7 mm) HD95 of gastric motion. Results showed that reducing of train time window (5–2 min) does not influence the prediction performance, while using 5 spokes increased prediction errors. Significance. The proposed gastric motion prediction model has sufficiently accurate prediction performance to allow for sub-millimeter accuracy while allowing sufficient time for data processing and machine interaction and shows the potential for clinical implementation to support stomach motion tracking during radiotherapy.

Predictive processing in depression: Increased prediction error following negative valence contexts and influence of recent mood-congruent yet irrelevant experiences

BackgroundNovel theoretical models of depression have recently emerged based on an influential new perspective in neuroscience known as predictive processing. In these models, depression may be understood as an imbalance of predictive signals in the brain; more specifically, a dominance of predictions leading to a relative insensitivity to prediction error. Despite these important theoretical advances, empirical evidence remains limited, and how expectations are generated and used dynamically in individuals with depression remains largely unexplored. MethodsIn this study, we induced facial expression predictions using emotion contexts in 34 individuals with depression and 34 healthy controls. ResultsCompared to controls, individuals with depression perceived displayed facial expressions as less similar to their expectations (i.e., increased difference between expectations and actual sensory input) following contexts evoking negative valence emotions, indicating that depressed individuals have increased prediction error in such contexts. This effect was amplified by recent mood-congruent yet irrelevant experiences. LimitationsThe clinical sample included participants with comorbid psychopathology and taking medication. Additionally, the two groups were not evaluated in the same setting, and only three emotion categories (fear, sadness, and happiness) were explored. ConclusionsOur results shed light on potential mechanisms underlying processing abnormalities regarding negative information, which has been consistently reported in depression, and may be a relevant point of departure for exploring transdiagnostic vulnerability to mental illness. Our data also has the potential to improve clinical practice through the implementation of novel diagnostic and therapeutic tools based on the assessment and modulation of predictive signals.

Influence of Subject-Specific Effects in Longitudinal Modelling of Cognitive Decline in Alzheimer's Disease.

Background:Accurate longitudinal modelling of cognitive decline is a major goal of Alzheimer’s disease and related dementia (ADRD) research. However, the impact of subject-specific effects is not well characterized and may have implications for data generation and prediction.Objective:This study seeks to address the impact of subject-specific effects, which are a less well-characterized aspect of ADRD cognitive decline, as measured by the Alzheimer’s Disease Assessment Scale’s Cognitive Subscale (ADAS-Cog).Methods:Prediction errors and biases for the ADAS-Cog subscale were evaluated when using only population-level effects, robust imputation of subject-specific effects using model covariances, and directly known individual-level effects fit during modelling as a natural control. Evaluated models included pre-specified parameterizations for clinical trial simulation, analogous mixed-effects regression models parameterized directly, and random forest ensemble models. Assessment used a meta-database of Alzheimer’s disease studies with validation in simulated synthetic cohorts.Results:All models observed increases in variance under imputation leading to increased prediction error. Bias decreased with imputation except under the pre-specified parameterization, which increased in the meta-database, but was attenuated under simulation. Known fitted subject effects gave the best prediction results.Conclusion:Subject-specific effects were found to have a profound impact on predicting ADAS-Cog. Reductions in bias suggest imputing random effects assists in calculating results on average, as when simulating clinical trials. However, reduction in error emphasizes population-level effects when attempting to predict outcomes for individuals. Forecasting future observations greatly benefits from using known subject-specific effects.

Alternatives to titanium dioxide in tablet coating

Titanium dioxide (TiO2) is one of the most commonly used pharmaceutical excipients. It is widely used as a white pigment in tablet and pellet coatings. However, it has recently been under massive criticism as a number of studies suggest a cancerogenic potential. It can therefore no longer be taken for granted that TiO2 will continue to be universally available for drug products. Finding suitable alternatives is hence of special relevance. In this study, a number of different pigments were coated on tablets and their covering potential analyzed. None of the alternative pigments showed comparable effectiveness and efficiency to TiO2, though the CaCO3/CaHPO4-based coating showed the second-best results. Regarding the ability to protect photosensitive active ingredients, ZnO showed a comparable potential as TiO2, while all other pigments failed. Using the alternative pigments as markers for in-line Raman spectroscopy as a process analytical technology was challenging and led to increased prediction errors. Again, the CaCO3/CaHPO4-based coating was the only of the tested alternatives with satisfying results, while all other pigments led to unacceptably high prediction errors.

Individual differences in sensory and expectation driven interoceptive processes: a novel paradigm with implications for alexithymia, disordered eating and obesity

Those with disordered eating and/or obesity often express difficulties in sensing or interpreting what is happening in the body (interoception). However, research is hindered by conceptual confusion, concerns surrounding domain specificity, and an inability to distinguish sensory (bottom-up) and expectation driven (top-down) interoceptive processes. A paradigm was therefore developed from an active inference perspective. Novel indices were computed and examined in those with alexithymia: a personality associated with interoceptive deficits and disordered eating. The paradigm successfully identified individuals driven by sensations rather than expectations: alexithymia was characterized by attenuated prior precision (a larger divergence between pre-prandial and post-prandial satiety, and low expectation confidence), and increased prediction error (a higher correlation between changes in hunger and blood glucose, and greater rebound hunger after a sensory incongruent drink). In addition, those with a higher BMI were less confident and had a larger anticipated satiety divergence. These findings demonstrate the need to move beyond existing paradigms such as the Satiety Quotient and Heartbeat Counting Task which may have limited our understanding of eating behaviour.

Systematic evaluation of supervised machine learning for sample origin prediction using metagenomic sequencing data

BackgroundThe advent of metagenomic sequencing provides microbial abundance patterns that can be leveraged for sample origin prediction. Supervised machine learning classification approaches have been reported to predict sample origin accurately when the origin has been previously sampled. Using metagenomic datasets provided by the 2019 CAMDA challenge, we evaluated the influence of variable technical, analytical and machine learning approaches for result interpretation and novel source prediction.ResultsComparison between 16S rRNA amplicon and shotgun sequencing approaches as well as metagenomic analytical tools showed differences in normalized microbial abundance, especially for organisms present at low abundance. Shotgun sequence data analyzed using Kraken2 and Bracken, for taxonomic annotation, had higher detection sensitivity. As classification models are limited to labeling pre-trained origins, we took an alternative approach using Lasso-regularized multivariate regression to predict geographic coordinates for comparison. In both models, the prediction errors were much higher in Leave-1-city-out than in 10-fold cross validation, of which the former realistically forecasted the increased difficulty in accurately predicting samples from new origins. This challenge was further confirmed when applying the model to a set of samples obtained from new origins. Overall, the prediction performance of the regression and classification models, as measured by mean squared error, were comparable on mystery samples. Due to higher prediction error rates for samples from new origins, we provided an additional strategy based on prediction ambiguity to infer whether a sample is from a new origin. Lastly, we report increased prediction error when data from different sequencing protocols were included as training data.ConclusionsHerein, we highlight the capacity of predicting sample origin accurately with pre-trained origins and the challenge of predicting new origins through both regression and classification models. Overall, this work provides a summary of the impact of sequencing technique, protocol, taxonomic analytical approaches, and machine learning approaches on the use of metagenomics for prediction of sample origin.

O9.5. NORMALIZATION OF DISTURBANCES IN PREDICTION ERROR IS RELATED TO TREATMENT RESPONSE AND RELATED TO THALAMIC GLUTAMATE LEVELS IN NON-RESPONDERS

BackgroundPrediction error is the mismatch between expected and obtained outcome, and psychosis has been linked to aberrant striatal prediction error signal. Several lines of evidence indicate alterations of the glutamatergic system to be involved in the pathophysiology of schizophrenia. We have previously reported abnormal thalamic glutamate levels at illness onset in schizophrenia patients driven by increased levels in non-responding patients, and that glutamatergic levels in the thalamus in twins dis- or concordant for psychosis were heritable and associated with the illness. Glutamatergic abnormalities may affect processing of prediction error; however, it remains unresolved if prediction error is affected by antipsychotic treatment, and to which extend treatment effect on prediction error is predicted by glutamatergic levels in patients characterized as responders or non-responders.Here, we explore treatment effects of aripiprazole on striatal prediction error signal in initially antipsychotic-naïve patients characterized as responders and non-responders and relate the findings to thalamic glutamate levels. We hypothesize a different treatment response in prediction error signal in responders and non-responders, and an association to baseline glutamate levels.MethodsThirty-three patients (age 22 ± 4 years) and 33 healthy controls (HC) matched on age and gender underwent functional Magnetic Resonance imaging (fMRI) and magnetic resonance spectroscopy (1H-MRS) (3T) at baseline and after 6 weeks of treatment with aripiprazole.Prediction error related brain activity was examined using a Monetary Incentive Delay Task. Glutamate levels were estimated in the left thalamus and analyzed using LCModel. In patients, symptom severity was assessed with the Positive and Negative Syndrome Scale. The Andreasen criteria defined responders (N=10) and non- responders (N=23).Repeated measures analysis of variance was used to test the effect of time in prediction error signal with group (responders vs non-responders vs HC) as between subject factor and time as within factor. Analysis of variance, two sample t-test and paired sample t-test evaluated group differences at baseline and follow up. In a multiple regression analyses we investigated the influence of baseline glutamate levels, symptom severity and p-aripiprazole on changes in prediction error signal in both responders and non-responders.ResultsRepeated measures analysis of prediction error showed an effect of group (p=0.007) and no effect of time (p=0.29) or interaction (p=0.29). The effect of group was explained by an abnormal increased prediction error signal in responders (p=0.047) and non-responders (p=0.011) compared to HC at baseline, which was normalized at follow up in responders (p=0.94) but not in non-responders (p=0.02) compared to HC. Changes in prediction error signal following treatment were predicted by glutamate levels in non-responders (p=0.03) but not in responders (p=0.85) whereas p-aripiprazole and symptom severity did not predict changes in prediction error signal (all p>0.05).DiscussionThe findings suggest that treatment with a partial dopamine agonist normalizes prediction error signal in patients characterized as responders. Thalamic glutamate seems to play a role in the neural coding of prediction error in patients characterized as non-responders, where increased levels of glutamate in the thalamus seems to predict a less pronounced changes in prediction error signal.

M19. NIGROSTRIATAL CONNECTIVITY AND THE PREDICTION OF THOUGHT DISTURBANCE IN EARLY PSYCHOSIS

BackgroundDopamine neurons are known to fire both tonically and phasically, resulting in tonic dopamine concentrations and spikes in those concentrations (often referred to as transients). Empirical evidence has shown elevated activity in the striatum in response to neutral stimuli which correlated with positive symptoms, in line with the proposed increased prediction errors. The increase of sponataneous phasic dopamine release in early psychosis should also be evident by altered resting state connectivity between the midbrain and its dopaminergic projections to the dorsal striatum. Given that all antipsychotics bind to striatal dopamine D2 receptors, we assumed that this altered functional connectivity would also relate to treatment efficacy of those antipsychotic agents. Using striatal seeds, and extending our own group’s prior work exploring striatal connectivity, we computed an index of nigrostriatal connectivity and estimated its relationship with improvement of positive symptoms under antipsychotic monotherapy.MethodsWe conducted a parcellation of subcortical midbrain structures including the substantia nigra and ventral tegmental area in two independent early phase psychosis cohorts. Early phase psychosis was defined as having an antipsychotic life time exposure of less than 52 weeks. Patients were between 15 and 40 years old and with a DSM-V-defined diagnosis of schizophrenia, schizoaffective disorder, or bipolar disorder with psychotic features and provided informed consent to participate in the study. Exclusion criteria were active alcohol or substance abuse disorders, serious medical conditions, and pregnancy. Preprocessing involved the state-of-the art Human Connectome Project pipeline and extensive correction for motion artifacts. After parcellation, we extracted the connectivity values of the striatal seeds with the substantia nigra and calculated an improved version of the striatal connectivity index with the R-package sci developed by the presenter and freely available online at http://github.com/philipphoman/sci.ResultsWe found that nigrostriatal connectivity was predictive of treatment efficacy in a sample of 41 patients with early phase psychosis. Indeed, individual differences in this connectivity index explained 18% of the variance of positive symptom improvement under antipsychotic treatment. Importantly, we replicated this effect in an independent replication cohort of 40 early phase psychosis patients, where we found that the identical model explained 11% of the variance in symptom improvement.DiscussionThere is an urgent need for diagnostic and prognostic biomarkers in psychotic disorders, and our results suggest that nigrostriatal connectivity is a promising candidate for a prognostic biomarker for antipsychotic treatment.

Migration and schizophrenia: meta-analysis and explanatory framework.

Systematic reviews and meta-analyses suggest that there are increased rates of schizophrenia and related psychoses in first- and second-generation migrants and refugees. Here, we present a meta-analysis on the incidence of non-affective psychotic disorders among first- and second-generation migrants. We found substantial evidence for an increased relative risk of incidence among first- and second-generation migrants compared to the native population. As heterogeneity of included studies was high, effect estimates should be interpreted with caution and as guiding values rather than exact risk estimates. We interpret our findings in the context of social exclusion and isolation stress, and provide an explanatory framework that links cultural differences in verbal communication and experienced discrimination with the emergence of psychotic experiences and their neurobiological correlates. In this context, we discuss studies observing stress-dependent alterations of dopamine neurotransmission in studies among migrants versus non-migrants as well as in subjects with psychotic disorders. We suggest that social stress effects can impair contextualization of the meaning of verbal messages, which can be accounted for in Bayesian terms by a reduced precision of prior beliefs relative to sensory data, causing increased prediction errors and resulting in a shift towards the literal or "concrete" meaning of words. Compensatory alterations in higher-level beliefs, e.g., in the form of generalized interpretations of ambiguous interactions as hostile behavior, may contribute to psychotic experiences in migrants. We thus suggest that experienced discrimination and social exclusion is at the core of increased rates of psychotic experiences in subjects with a migration background.