Purpose of review

The field of neurotrauma research faces a reproducibility crisis. In response, research leaders in traumatic brain injury (TBI) and spinal cord injury (SCI) are leveraging data curation and analytics methods to encourage transparency, and improve the rigor and reproducibility. Here we review the current challenges and opportunities that come from efforts to transform neurotrauma's big data to knowledge.

Recent findings

Three parallel movements are driving data-driven-discovery in neurotrauma. First, large multicenter consortia are collecting large quantities of neurotrauma data, refining common data elements (CDEs) that can be used across studies. Investigators are now testing the validity of CDEs in diverse research settings. Second, data sharing initiatives are working to make neurotrauma data findable, accessible, interoperable, and reusable (FAIR). These efforts are reflected by recent open data repository projects for preclinical and clinical neurotrauma. Third, machine learning analytics are allowing researchers to uncover novel data-driven-hypotheses and test new therapeutics in multidimensional outcome space.

Summary

We are on the threshold of a new era in data collection, curation, and analysis. The next phase of big data in neurotrauma research will require responsible data stewardship, a culture of data-sharing, and the illumination of 'dark data'.

Translation of traumatic brain injury (TBI) research findings from bench to bedside involves aligning multi-species data across diverse data types including imaging and molecular biomarkers, histopathology, behavior, and functional outcomes. In this review we argue that TBI translation should be acknowledged for what it is: a problem of big data that can be addressed using modern data science approaches. We review the history of the term big data, tracing its origins in Internet technology as data that are "big" according to the "4Vs" of volume, velocity, variety, veracity and discuss how the term has transitioned into the mainstream of biomedical research. We argue that the problem of TBI translation fundamentally centers around data variety and that solutions to this problem can be found in modern machine learning and other cutting-edge analytical approaches. Throughout our discussion we highlight the need to pull data from diverse sources including unpublished data ("dark data") and "long-tail data" (small, specialty TBI datasets undergirding the published literature). We review a few early examples of published articles in both the pre-clinical and clinical TBI research literature to demonstrate how data reuse can drive new discoveries leading into translational therapies. Making TBI data resources more Findable, Accessible, Interoperable, and Reusable (FAIR) through better data stewardship has great potential to accelerate discovery and translation for the silent epidemic of TBI.

Meta-analyses suggest that the published literature represents only a small minority of the total data collected in biomedical research, with most becoming 'dark data' unreported in the literature. Dark data is due to publication bias toward novel results that confirm investigator hypotheses and omission of data that do not. Publication bias contributes to scientific irreproducibility and failures in bench-to-bedside translation. Sharing dark data by making it Findable, Accessible, Interoperable, and Reusable (FAIR) may reduce the burden of irreproducible science by increasing transparency and support data-driven discoveries beyond the lifecycle of the original study. We illustrate feasibility of dark data sharing by recovering original raw data from the Multicenter Animal Spinal Cord Injury Study (MASCIS), an NIH-funded multi-site preclinical drug trial conducted in the 1990s that tested efficacy of several therapies after a spinal cord injury (SCI). The original drug treatments did not produce clear positive results and MASCIS data were stored in boxes for more than two decades. The goal of the present study was to independently confirm published machine learning findings that perioperative blood pressure is a major predictor of SCI neuromotor outcome (Nielson et al., 2015). We recovered, digitized, and curated the data from 1125 rats from MASCIS. Analyses indicated that high perioperative blood pressure at the time of SCI is associated with poorer health and worse neuromotor outcomes in more severe SCI, whereas low perioperative blood pressure is associated with poorer health and worse neuromotor outcome in moderate SCI. These findings confirm and expand prior results that a narrow window of blood-pressure control optimizes outcome, and demonstrate the value of recovering dark data for assessing reproducibility of findings with implications for precision therapeutic approaches.

In general, glucose consumption improves cognitive performance; however, it is unknown whether glucose specifically alters attentional food bias, and how this process may vary by BMI status. We hypothesized that glucose consumption would increase attentional food bias among individuals of obese BMI status more so than among individuals of lean BMI status. Participants (N = 35) completed the n-back, a working memory task modified to assess attentional food bias (ATT-Food), under fasting and glucose challenge conditions. We computed pre-post changes in ATT-Food, blood glucose and insulin (∆BG & ∆BI), and perceived task-stress (∆stress). After the second cognitive test and blood draw, participants ate lunch and completed a "taste test" of highly palatable foods, and we recorded food consumption. Pre-post changes in ATT-Food were greater among participants of obese (relative to lean) BMI status (F(1,33) = 5.108, p = .031). Greater ∆ATT-Food was significantly associated with greater ∆BG (r = .462, p = .007) and reduced ∆stress (r =-.422, p = .011), and marginally associated with greater taste-test eating (r =.325, p = .057), but was not associated with ∆BI. Our findings suggest that individuals of obese BMI status may exhibit "sweet cognition," as indexed by greater attentional food bias following glucose ingestion, relative to individuals of lean BMI status. Among individuals of obese BMI status, sweet cognition may arise from difficulty broadening attention toward non-food cues after consuming a high glucose load, thereby potentially perpetuating sugar consumption. If confirmed by further research, measures of sweet cognition may help identify individuals with a phenotype of risk for obesity and greater sugar consumption, who may benefit from tailored interventions.

Background

Chronic psychological stress is a risk factor for cardiovascular disease and mortality. Circulating hematopoietic progenitor cells (CPCs) maintain vascular homeostasis, correlate with preclinical atherosclerosis, and prospectively predict cardiovascular events. We hypothesize that (1) chronic caregiving stress is related to reduced CPC number, and (2) this may be explained in part by negative interactions within the family.

Methods

We investigated levels of stress and CPCs in 68 healthy mothers - 31 of these had children with an autism spectrum disorder (M-ASD) and 37 had neurotypical children (M-NT). Participants provided fasting blood samples, and CD45⁺CD34⁺KDR⁺ and CD45⁺CD133⁺KDR⁺ CPCs were assayed by flow cytometry. We averaged the blom-transformed scores of both CPCs to create one index. Participants completed the perceived stress scale (PSS), the inventory for depressive symptoms (IDS), and reported on daily interactions with their children and partners, averaged over 7 nights.

Results

M-ASD exhibited lower CPCs than M-NT (Cohen's d=0.83; p⩽0.01), controlling for age, BMI, and physical activity. Across the whole sample, positive interactions were related to higher CPCs, and negative interactions to lower CPCs (allp's<0.05). The adverse effects of group on CPCs were significantly mediated through negative interactions with the child (indirect β=-0.24, p⩽0.01). In the full model, greater age (β=-0.19, p=0.04), BMI (β=-0.18, p=0.04), and negative interactions with the child (β=-0.33, p<0.01) were independently associated with lower CPCs. M-ASD had a less healthy lipid profile (total cholesterol/HDL), which in turn, was associated with lower CPCs.

Conclusions

Chronic stress adversely impacts CPC number, an early-stage biomarker that predicts subclinical atherosclerosis and future CVD events, independent of traditional cardiovascular risk factors and inflammatory factors. Among maternal caregivers, child-related interpersonal stress appears to be a key psychological predictor of stress-related CVD risk.

The past decade has seen accelerating movement from data protectionism in publishing toward open data sharing to improve reproducibility and translation of biomedical research. Developing data sharing infrastructures to meet these new demands remains a challenge. One model for data sharing involves simply attaching data, irrespective of its type, to publisher websites or general use repositories. However, some argue this creates a 'data dump' that does not promote the goals of making data Findable, Accessible, Interoperable and Reusable (FAIR). Specialized data sharing communities offer an alternative model where data are curated by domain experts to make it both open and FAIR. We report on our experiences developing one such data-sharing ecosystem focusing on 'long-tail' preclinical data, the Open Data Commons for Spinal Cord Injury (odc-sci.org). ODC-SCI was developed with community-based agile design requirements directly pulled from a series of workshops with multiple stakeholders (researchers, consumers, non-profit funders, governmental agencies, journals, and industry members). ODC-SCI focuses on heterogeneous tabular data collected by preclinical researchers including bio-behaviour, histopathology findings and molecular endpoints. This has led to an example of a specialized neurocommons that is well-embraced by the community it aims to serve. In the present paper, we provide a review of the community-based design template and describe the adoption by the community including a high-level review of current data assets, publicly released datasets, and web analytics. Although odc-sci.org is in its late beta stage of development, it represents a successful example of a specialized data commons that may serve as a model for other fields.

Background

Predicting neurological recovery after spinal cord injury (SCI) is challenging. Using topological data analysis, we have previously shown that mean arterial pressure (MAP) during SCI surgery predicts long-term functional recovery in rodent models, motivating the present multicenter study in patients.

Methods

Intra-operative monitoring records and neurological outcome data were extracted (n = 118 patients). We built a similarity network of patients from a low-dimensional space embedded using a non-linear algorithm, Isomap, and ensured topological extraction using persistent homology metrics. Confirmatory analysis was conducted through regression methods.

Results

Network analysis suggested that time outside of an optimum MAP range (hypotension or hypertension) during surgery was associated with lower likelihood of neurological recovery at hospital discharge. Logistic and LASSO (least absolute shrinkage and selection operator) regression confirmed these findings, revealing an optimal MAP range of 76-[104-117] mmHg associated with neurological recovery.

Conclusions

We show that deviation from this optimal MAP range during SCI surgery predicts lower probability of neurological recovery and suggest new targets for therapeutic intervention.

Funding

NIH/NINDS: R01NS088475 (ARF); R01NS122888 (ARF); UH3NS106899 (ARF); Department of Veterans Affairs: 1I01RX002245 (ARF), I01RX002787 (ARF); Wings for Life Foundation (ATE, ARF); Craig H. Neilsen Foundation (ARF); and DOD: SC150198 (MSB); SC190233 (MSB).