Background Vocational aptitude and cognitive resilience predict military success, yet current assessments rely on resource-intensive, in-person testing that limits scalability. A brief, self-administered, remotely deployable computerized battery offers a practical solution for large-scale screening and monitoring. Objective This study aims to deploy a set of computerized assessments among National Guard recruits and assess their preliminary construct validity against a standardized aptitude measure and a research-based proxy for cognitive resilience. Methods In this observational study, 267 enlisted service members from the Minnesota Army National Guard participated in 2 complementary ethics-approved observational trials: Office of Naval Research Neuropsychometrics and Advancing Research on Mechanisms of Resilience (ARMOR). National Guard soldiers in ARMOR completed the Armed Forces Qualification Test (AFQT), Penn Computerized Neurocognitive Battery (Penn CNB), and a 20-minute computerized brain health assessment battery (BrainHQ) at separate time points over the course of their military careers. BrainHQ assessments consisted of adaptive psychophysical tasks measuring the speed and accuracy of visual and auditory information processing. The battery assessed decision-making speed, emotion-processing speed, selective attention under speeded conditions, working memory capacity for speeded visual elements, verbal memory and learning of speeded speech, and problem-solving speed. The Penn CNB included nonspeeded neuropsychological assessments of executive function, verbal memory, social cognition, and reasoning. Linear regression evaluated the association between BrainHQ performance and AFQT percentiles, and partial correlations assessed associations between conceptually related BrainHQ and Penn CNB subtests. Results Participants were predominantly young (mean age 19.1 years) and male (178/267, 66.7%). BrainHQ performance was significantly associated with enlistment eligibility and vocational aptitude, as measured by the AFQT (P<.001), after controlling for age and education. The overall model explained 24.4% of the variance in AFQT percentiles (adjusted R2=0.227). The BrainHQ assessment composite was the strongest predictor, uniquely accounting for 19.2% of the variance and supporting the construct validity of aptitude. These associations persisted despite the temporal separation between assessment time points. Quartile analyses showed graded relationships between BrainHQ performance and AFQT eligibility thresholds, with higher BrainHQ performance associated with progressively greater probabilities of meeting higher AFQT benchmarks. Preplanned partial correlations between BrainHQ subtests and standardized neurocognitive measures from the Penn CNB showed significant positive associations (r=0.17-0.25; all P<.001 to .02) with cognitive domains typically associated with cognitive resilience. Conclusions A brief, self-administered, and scalable brain health battery demonstrates associations with military vocational aptitude and with neurocognitive domains associated with cognitive resilience. Future studies should evaluate whether integrating these assessments into current practices predicts success in Basic Combat Training, guides military progression, and supports long-term cognitive screening and monitoring across the Armed Forces.

Study Objectives:Shifts in the light-dark cycle (L:D cycle) often trigger phase shifts in physiological data related to the sleep-wake cycle. Slow wave activity (delta) indicates sleep pressure and intensity. This study examines how delta power adapts to shifts in L:D cycle and the temporal dynamics of its coupling with rest-activity rhythms during re-entrainment. Methods:We collected electroencephalogram (EEG) and accelerometer data from three non-human primates during baseline and shifted (8-h delayed light-on) conditions. We derived delta power (0.5 ~ 4 Hz) using Fast Fourier Transform. To quantify changes in delta power dynamics following L:D cycle shifts, we calculated diurnal differences in delta power, per cent variance explained by time-of-day, circadian coupling with physical activity, and delta power activity transitions timing. Results:In both conditions, delta power exhibited a robust 24-h periodicity, and a significant portion of the variance (57.61% ± 6.99%) could be explained by time of day. We found an early transition of delta power in the first 2 days of the shifted condition, followed by realignment to the light-off time within 3 days after the shift. We used coherence analysis to reveal strong coupling between delta power and locomotor activity, with a consistent anti-phase relationship across baseline and phase shifted conditions. Conclusions:Our findings demonstrate that delta power adapts rapidly to environmental phase shifts while maintaining circadian rhythmicity and stable coordination with rest-activity rhythms. Here, we provide new insight into how neural and behavioral states remain aligned during circadian disruptions in a diurnal species.

Adaptive decision-making requires balancing exploitation of known rewarding options with exploration of uncertain alternatives, a dilemma also known as the exploration-exploitation tradeoff. While this framework has been widely studied in reinforcement learning research, its relevance to coping, defined as the cognitive and behavioral strategies that individuals use to manage stress and uncertainty, remains underexplored. Maladaptive coping may reflect rigidity in exploitation or ineffective exploration, whereas adaptive coping may involve flexible adjustment of control in changing environments. In this study, we examined whether interindividual differences in the transition dynamics between explore and exploit strategy predict coping styles in a large online general population sample. A total of 1732 participants completed a three-armed restless bandit task, and their latent explore-exploit strategy states and transition patterns were modeled using a Hidden Markov Model. These computational indices of explore-exploit dynamics were then linked to self-reported psychological coping strategies using regression and canonical correlation analysis. Individuals with a greater tendency to persist in exploitative states reported less reliance on avoidant and emotion-focused coping, whereas exploratory tendencies showed distinct associations with externally oriented coping strategies. Unsupervised clustering of exploration-exploitation dynamics further revealed four distinct decision-making subtypes, each associated with unique coping profiles. These findings provide the first evidence that computational markers of explore-exploit control dynamics relate to psychological coping profiles, offering mechanistic insight into psychological adaptation and resilience.

BackgroundImpairments in cognition and motivation are core features of psychosis and strong predictors of social and occupational functioning. Accumulating evidence indicates that cognitive deficits in psychosis can be improved by computer-based cognitive training programs; however, barriers include access and adherence to cognitive training exercises. Limited evidence-based methods have been established to enhance motivated behavior. In this study, we tested the effects of web-based targeted cognitive and social cognitive training (TCT) delivered in conjunction with an innovative digital smartphone app called Personalized Real-Time Intervention for Motivational Enhancement (PRIME). The PRIME app provides users with a motivational coach to set personalized goals and secure social networking for peer support. ObjectiveThis study investigated whether deficits in cognition and motivation in people with a psychosis spectrum disorder (N=100) can be successfully addressed with 30 hours of TCT+PRIME as compared with 30 hours of a computer games control condition (CG) plus PRIME (CG+PRIME). Here, we describe our study procedures, the feasibility and acceptability of the intervention, and the results on all primary outcomes. MethodsIn this double-blind randomized controlled trial, English-speaking participants completed all cognitive training, PRIME activities, and assessments remotely. Participants completed a diagnostic interview and remote cognitive, clinical, and self-report measures at baseline, posttraining, and at a 6-month follow-up. ResultsThis study included participants from 27 states across the United States and 8 countries worldwide. The study population was 58% (58/100) female, with a mean age of 33.77 (SD 10.70) years. On average, participants completed more than half of the cognitive training regimen (mean 18.58, SD 12.47 hours of training), and logged into the PRIME app 4.71 (SD 1.58) times per week. The attrition rate of 22% (22/100) was lower than that reported in our previous studies on remote cognitive training. The total sample showed significant gains in global cognition (P=.03) and attention (P<.001). The TCT+PRIME participants showed significantly greater gains in emotion recognition (P<.001) and global cognition at the trend level (P=.09), although this was not statistically significant, relative to the CG+PRIME participants. The total sample also showed significant improvements on multiple indices of motivation (P=.02-0.05), in depression (P=.04), in positive symptoms (P=.04), and in negative symptoms at a trend level (P=.09), although this was not statistically significant. Satisfaction with the PRIME app was rated at 7.74 (SD 2.05) on a scale of 1 to 10, with higher values indicating more satisfaction. ConclusionsThese results demonstrate the feasibility and acceptability of remote cognitive training combined with the PRIME app and that this intervention can improve cognition, motivation, and symptoms in individuals with psychosis. TCT+PRIME appeared more effective in improving emotion recognition and global cognition than CG+PRIME. Future analyses will test the relationship between hours of cognitive training completed; PRIME use; and changes in cognition, motivation, symptoms, and functioning. Trial RegistrationClinicalTrials.gov NCT02782442; https://clinicaltrials.gov/study/NCT02782442

Impairments in cognition and motivation are core features of psychosis and strong predictors of social and occupational functioning. Accumulating evidence indicates cognitive deficits in psychosis can be improved by computer-based cognitive training programs; however, barriers include access and adherence to cognitive training exercises and limited evidence-based methods to enhance motivated behavior. In this study, we tested the effects of online targeted cognitive and social cognitive training (TCT), delivered in conjunction with an innovative digital smartphone app called Personalized Real-Time Intervention for Motivational Enhancement (PRIME). The PRIME app provides users with a motivation coach to set personalized goals and secure social networking for peer support. We investigated whether deficits in cognition and motivation in people with a psychosis spectrum illness can be successfully addressed by comparing the effects of 30 hours of TCT+PRIME to 30 hours of a computer games control condition plus PRIME (CG+PRIME). Here, we describe our study procedures; the features of the interventions; our recruitment of individuals across the United States and internationally; and the feasibility and acceptability of the intervention in 98 randomized participants. In this double-blind, randomized, controlled trial (ClinicalTrials.gov Identifier: NCT02782442), English-speaking participants completed all cognitive training, PRIME activities, and assessments remotely. Participants completed a diagnostic interview and remote cognitive, clinical, and self-report measures at baseline, post training, and at a six-month follow-up. Our feasibility results include participants from 27 states across the U.S. and 8 countries worldwide. Our study population was 59% female and a mean age of 33.87 (SD=10.75). On average, participants completed more than half of the cognitive training regimen (M=19.12, SD=12.38 hours of training), logged into the PRIME application 4.72 (SD=1.57) times per week, interacted with their PRIME coach 78.47 (SD=89.05) times during the 16 week intervention, and achieved 14.03 (SD=14.63) goals. The attrition rate of 20% was lower than our previous studies of remote cognitive training (30%). Hours of cognitive training or computer games completed was significantly associated with the number of PRIME logins, coach interactions, and goals achieved. In addition, the number of goals achieved was significantly associated with the number of logins, and with coach and peer interactions. Overall satisfaction with the PRIME app was rated at 7.74 (SD=2.05) on a scale of 1-10, with higher values indicating more satisfaction. These results demonstrate the feasibility and acceptability of remote cognitive training combined with the PRIME application. Importantly, more hours of cognitive training were completed and a greater number of goals were achieved in participants who engaged more with the app. These results and our lower attrition rate suggest that PRIME may improve motivation and adherence with cognitive training. Future analyses will test for group differences in changes in cognition, symptoms, and functioning. ClinicalTrials.gov Identifier: NCT02782442

Affective behaviours and mental health are profoundly affected by disturbances in circadian rhythms. Casein kinase 1 epsilon (CSNK1E) is a core component of the circadian clock. Mice with tau or null mutation of this gene have shortened and lengthened circadian period respectively. Here, we examined anxiety-like, fear, and despair behaviours in both male and female mice of these two different mutants. Compared with wild-type mice, we found reductions in fear and anxiety-like behaviours in both mutant lines and in both sexes, with the tau mutants exhibiting the greatest phenotypic changes. However, the behavioural despair had distinct phenotypic patterns, with markedly less behavioural despair in female null mutants, but not in tau mutants of either sex. To determine whether abnormal light entrainment of tau mutants to 24-h light-dark cycles contributes to these phenotypic differences, we also examined these behaviours in tau mutants on a 20-h light-dark cycle close to their endogenous circadian period. The normalized entrainment restored more wild-type-like behaviours for fear and anxiety, but it induced behavioural despair in tau mutant females. These data show that both mutations of Csnk1e broadly affect fear and anxiety-like behaviours, while the effects on behavioural despair vary with genetics, photoperiod, and sex, suggesting that the mechanisms by which Csnk1e affects fear and anxiety-like behaviours may be similar, but distinct from those affecting behavioural despair. Our study also provides experimental evidence in support of the hypothesis of beneficial outcomes from properly entrained circadian rhythms in terms of the anxiety-like and fear behaviours.

Background In recent years, there has been increased interest in the development of remote psychological assessments. These platforms increase accessibility and allow clinicians to monitor important health metrics, thereby informing patient-centered treatment. Objective In this study, we report the properties and usability of a new web-based neurocognitive assessment battery and present a normative data set for future use. Methods A total of 781 participants completed a portion of 8 tasks that captured performance in auditory processing, visual-spatial working memory, visual-spatial learning, cognitive flexibility, and emotional processing. A subset of individuals (n=195) completed a 5-question survey measuring the acceptability of the tasks. Results Between 252 and 426 participants completed each task. Younger individuals outperformed their older counterparts in 6 of the 8 tasks. Therefore, central tendency data metrics were presented using 7 different age bins. The broad majority of participants found the tasks interesting and enjoyable and endorsed some interest in playing them at home. Only 1 of 195 individuals endorsed not at all for the statement, “I understood the instructions.” Older individuals were less likely to understand the instructions; however, 72% (49/68) of individuals over the age of 60 years still felt that they mostly or very much understood the instructions. Conclusions Overall, the tasks were found to be widely acceptable to the participants. The use of web-based neurocognitive tasks such as these may increase the ability to deploy precise data-informed interventions to a wider population.

Affective behaviors and mental health are profoundly affected by disturbances in circadian rhythms. Casein kinase 1 epsilon (CSNK1E) is an essential component of the core circadian clock. Mice with tau or null mutation of this gene have shortened and lengthened circadian period respectively. Here we examined anxiety-like, fear, and depressive-like behaviors in both male and female mice of these two different mutants. Compared with wild-type mice, we found reductions in fear and anxiety-like behaviors in both mutant lines and in both sexes, with the tau mutants exhibiting the greatest phenotypic changes. However, the depressive-like behaviors had distinct phenotypic patterns, with markedly less depressive-like behaviors in female null mutants, but not in tau mutants of either sex. To determine whether abnormal light entrainment of tau mutants to 24 hour light-dark cycles contributes to these phenotypic differences, we also examined these behaviors in tau mutants on a 20 hour light-dark cycle close to their endogenous circadian period. The normalized entrainment restored more wild-type-like behaviors for fear and anxiety, but it induced depressive-like behavior in tau mutant females. These data show that both mutations of Csnk1e broadly affect fear and anxiety-like behaviors, while the effects on depressive-like behavior vary with genetics, photoperiod, and sex, suggesting that the mechanisms by which Csnk1e affects fear and anxiety-like behaviors may be similar, but distinct from those affecting depressive-like behavior. Our study also provides experimental evidence in support of the hypothesis of beneficial outcomes from properly entrained circadian rhythms in terms of the anxiety-like and fear behaviors.### Competing Interest StatementThe authors have declared no competing interest.* ANOVA : Analysis of Variance CSNK1D : Casein kinase 1 delta CSNK1E : Casein kinase 1 epsilon EPM : Elevated Plus Maze FCT : Fear Conditioning Test GSK3B : Glycogen synthase kinase 3 beta LD : Light-Dark OFA : Open Field Activity PCA : Principal Component Analysis TST : Tail Suspension Test ZT : Zeitgeber Time

To understand the transcriptomic organization underlying sleep and affective function, we studied a population of (C57BL/6J × 129S1/SvImJ) F2 mice by measuring 283 affective and sleep phenotypes and profiling gene expression across four brain regions. We identified converging molecular bases for sleep and affective phenotypes at both the single-gene and gene-network levels. Using publicly available transcriptomic datasets collected from sleep-deprived mice and patients with major depressive disorder (MDD), we identified three cortical gene networks altered by the sleep/wake state and depression. The network-level actions of sleep loss and depression were opposite to each other, providing a mechanistic basis for the sleep disruptions commonly observed in depression, as well as the reported acute antidepressant effects of sleep deprivation. We highlight one particular network composed of circadian rhythm regulators and neuronal activity-dependent immediate-early genes. The key upstream driver of this network, Arc, may act as a nexus linking sleep and depression. Our data provide mechanistic insights into the role of sleep in affective function and MDD.

The timing and propensity of sleep are regulated by two interactive processes: circadian rhythmicity and sleep homeostasis. It has been demonstrated that the two processes are likely to converge at molecular levels, involving networks of genes including the molecular circadian clock machinery. However, since the molecular pathways involved in sleep homeostasis are still elusive, it is unclear how the circadian and homeostatic signals integrate to regulate sleep beyond a handful of clock genes. In ~200 (C57BL/6J x 129S1/SvImJ) F2 mice, we collected a comprehensive dataset containing: 1) phenotypic data of affective behaviors and sleep (including baseline, recovery after 6-h sleep deprivation, and sleep after restraint stress), 2) genotypic data across the genome, and 3) microarray data in the prefrontal cortex, hippocampus, midbrain-thalamus, and hypothalamus. We reconstructed gene networks in each of these brain regions. We also integrated this dataset with multiple circadian and sleep genomics datasets that are publicly available, in order to identify convergent networks that link both circadian and sleep homeostatic processes. Using our mouse dataset, we uncovered brain-region-conserved, as well as brain-region-specific, gene networks that are associated with sleep phenotypes. These sleep gene networks are cell-type specific and can be functionally annotated with specific cellular processes, revealing molecular pathways key to sleep function and regulation. Via Integrated analysis with publicly available datasets, we identified a number of gene networks that are enriched with differentially expressed genes responding to sleep loss and cycling genes peaking at particular times of the day. Particularly, our analysis in the prefrontal cortex highlights a Clock-driven network and a sphingolipid-metabolism network, whose overall network gene expression are cycling with opposite phases and are affected by sleep deprivation in opposite directions. This analysis of molecular networks important to sleep regulation reveals novel insights into convergent pathways that integrate circadian timing and homeostatic signals in the regulation of sleep. This work was supported by a research grant from Merck Research Laboratories, by the Defense Advanced Research Projects Agency and the U.S. Army Research Office (W911NF101006), and by the National Institute of Mental Health of the NIH (J.R.S.; F30MH106293).

BACKGROUND:Chronic alcohol exposure exerts numerous adverse effects, although the specific mechanisms underlying these negative effects on different tissues are not completely understood. Alcohol also affects core properties of the circadian clock system, and it has been shown that disruption of circadian rhythms confers vulnerability to alcohol-induced pathology of the gastrointestinal barrier and liver. Despite these findings, little is known of the molecular interactions between alcohol and the circadian clock system, especially regarding implications for tissue-specific susceptibility to alcohol pathologies. The aim of this study was to identify changes in expression of genes relevant to alcohol pathologies and circadian clock function in different tissues in response to chronic alcohol intake.METHODS:Wild-type and circadian Clock(Δ19) mutant mice were subjected to a 10-week chronic alcohol protocol, after which hippocampal, liver, and proximal colon tissues were harvested for gene expression analysis using a custom-designed multiplex magnetic bead hybridization assay that provided quantitative assessment of 80 mRNA targets of interest, including 5 housekeeping genes and a predetermined set of 75 genes relevant for alcohol pathology and circadian clock function.RESULTS:Significant alterations in expression levels attributable to genotype, alcohol, and/or a genotype by alcohol interaction were observed in all 3 tissues, with distinct patterns of expression changes observed in each. Of particular interest was the finding that a high proportion of genes involved in inflammation and metabolism on the array was significantly affected by alcohol and the Clock(Δ19) mutation in the hippocampus, suggesting a suite of molecular changes that may contribute to pathological change.CONCLUSIONS:These results reveal the tissue-specific nature of gene expression responses to chronic alcohol exposure and the Clock(Δ19) mutation and identify specific expression profiles that may contribute to tissue-specific vulnerability to alcohol-induced injury in the brain, colon, and liver.

Sleep dysfunction and stress susceptibility are comorbid complex traits that often precede and predispose patients to a variety of neuropsychiatric diseases. Here, we demonstrate multilevel organizations of genetic landscape, candidate genes, and molecular networks associated with 328 stress and sleep traits in a chronically stressed population of 338 (C57BL/6J × A/J) F2 mice. We constructed striatal gene co-expression networks, revealing functionally and cell-type-specific gene co-regulations important for stress and sleep. Using a composite ranking system, we identified network modules most relevant for 15 independent phenotypic categories, highlighting a mitochondria/synaptic module that links sleep and stress. The key network regulators of this module are overrepresented with genes implicated in neuropsychiatric diseases. Our work suggests that the interplay among sleep, stress, and neuropathology emerges from genetic influences on gene expression and their collective organization through complex molecular networks, providing a framework for interrogating the mechanisms underlying sleep, stress susceptibility, and related neuropsychiatric disorders.

STUDY OBJECTIVE Sleep and mood disorders have long been understood to have strong genetic components, and there is considerable comorbidity of sleep abnormalities and mood disorders, suggesting the involvement of common genetic pathways. Here, we examine a candidate gene implicated in the regulation of both sleep and affective behavior using a knockout mouse model. DESIGN Previously, we identified a quantitative trait locus (QTL) for REM sleep amount, REM sleep bout number, and wake amount in a genetically segregating population of mice. Here, we show that traits mapping to this QTL correlated with an expression QTL for neurotensin receptor 1 (Ntsr1), a receptor for neurotensin, a ligand known to be involved in several psychiatric disorders. We examined sleep as well as behaviors indicative of anxiety and depression in the NTSR1 knockout mouse. MEASUREMENTS AND RESULTS NTSR1 knockouts had a lower percentage of sleep time spent in REM sleep in the dark phase and a larger diurnal variation in REM sleep duration than wild types under baseline conditions. Following sleep deprivation, NTSR1 knockouts exhibited more wake and less NREM rebound sleep. NTSR1 knockouts also showed increased anxious and despair behaviors. CONCLUSIONS Here we illustrate a link between expression of the Ntsr1 gene and sleep traits previously associated with a particular QTL. We also demonstrate a relationship between Ntsr1 and anxiety and despair behaviors. Given the considerable evidence that anxiety and depression are closely linked with abnormalities in sleep, the data presented here provide further evidence that neurotensin and Ntsr1 may be a component of a pathway involved in both sleep and mood disorders.

Despite the substantial impact of sleep disturbances on human health and the many years of study dedicated to understanding sleep pathologies, the underlying genetic mechanisms that govern sleep and wake largely remain unknown. Recently, the authors completed large-scale genetic and gene expression analyses in a segregating inbred mouse cross and identified candidate causal genes that regulate the mammalian sleep-wake cycle, across multiple traits including total sleep time, amounts of rapid eye movement (REM), non-REM, sleep bout duration, and sleep fragmentation. Here the authors describe a novel approach toward validating candidate causal genes, while also identifying potential targets for sleep-related indications. Select small-molecule antagonists and agonists were used to interrogate candidate causal gene function in rodent sleep polysomnography assays to determine impact on overall sleep architecture and to evaluate alignment with associated sleep-wake traits. Significant effects on sleep architecture were observed in validation studies using compounds targeting the muscarinic acetylcholine receptor M3 subunit (Chrm3) (wake promotion), nicotinic acetylcholine receptor alpha4 subunit (Chrna4) (wake promotion), dopamine receptor D5 subunit (Drd5) (sleep induction), serotonin 1D receptor (Htr1d) (altered REM fragmentation), glucagon-like peptide-1 receptor (Glp1r) (light sleep promotion and reduction of deep sleep), and calcium channel, voltage-dependent, T type, alpha 1I subunit (Cacna1i) (increased bout duration of slow wave sleep). Taken together, these results show the complexity of genetic components that regulate sleep-wake traits and highlight the importance of evaluating this complex behavior at a systems level. Pharmacological validation of genetically identified putative targets provides a rapid alternative to generating knock out or transgenic animal models, and may ultimately lead towards new therapeutic opportunities.

STUDY OBJECTIVE:Sleep-wake traits are well-known to be under substantial genetic control, but the specific genes and gene networks underlying primary sleep-wake traits have largely eluded identification using conventional approaches, especially in mammals. Thus, the aim of this study was to use systems genetics and statistical approaches to uncover the genetic networks underlying 2 primary sleep traits in the mouse: 24-h duration of REM sleep and wake. DESIGN:Genome-wide RNA expression data from 3 tissues (anterior cortex, hypothalamus, thalamus/midbrain) were used in conjunction with high-density genotyping to identify candidate causal genes and networks mediating the effects of 2 QTL regulating the 24-h duration of REM sleep and one regulating the 24-h duration of wake. SETTING:Basic sleep research laboratory. PATIENTS OR PARTICIPANTS:Male [C57BL/6J × (BALB/cByJ × C57BL/6J*) F1] N(2) mice (n = 283). INTERVENTIONS:None. MEASUREMENTS AND RESULTS:The genetic variation of a mouse N2 mapping cross was leveraged against sleep-state phenotypic variation as well as quantitative gene expression measurement in key brain regions using integrative genomics approaches to uncover multiple causal sleep-state regulatory genes, including several surprising novel candidates, which interact as components of networks that modulate REM sleep and wake. In particular, it was discovered that a core network module, consisting of 20 genes, involved in the regulation of REM sleep duration is conserved across the cortex, hypothalamus, and thalamus. A novel application of a formal causal inference test was also used to identify those genes directly regulating sleep via control of expression. CONCLUSION:Systems genetics approaches reveal novel candidate genes, complex networks and specific transcriptional regulators of REM sleep and wake duration in mammals.