Ian Barnett

Detecting Rare Variant Effects Using Extreme Phenotype Sampling in Sequencing Association Studies

In the increasing number of sequencing studies aimed at identifying rare variants associated with complex traits, the power of the test can be improved by guided sampling procedures. We confirm both analytically and numerically that sampling individuals with extreme phenotypes can enrich the presence of causal rare variants and can therefore lead to an increase in power compared to random sampling. Although application of traditional rare variant association tests to these extreme phenotype samples requires dichotomizing the continuous phenotypes before analysis, the dichotomization procedure can decrease the power by reducing the information in the phenotypes. To avoid this, we propose a novel statistical method based on the optimal Sequence Kernel Association Test that allows us to test for rare variant effects using continuous phenotypes in the analysis of extreme phenotype samples. The increase in power of this method is demonstrated through simulation of a wide range of scenarios as well as in the triglyceride data of the Dallas Heart Study.

Change Point Detection in Correlation Networks.

Many systems of interacting elements can be conceptualized as networks, where network nodes represent the elements and network ties represent interactions between the elements. In systems where the underlying network evolves, it is useful to determine the points in time where the network structure changes significantly as these may correspond to functional change points. We propose a method for detecting change points in correlation networks that, unlike previous change point detection methods designed for time series data, requires minimal distributional assumptions. We investigate the difficulty of change point detection near the boundaries of the time series in correlation networks and study the power of our method and competing methods through simulation. We also show the generalizable nature of the method by applying it to stock price data as well as fMRI data.

The Generalized Higher Criticism for Testing SNP-Set Effects in Genetic Association Studies

ABSTRACT It is of substantial interest to study the effects of genes, genetic pathways, and networks on the risk of complex diseases. These genetic constructs each contain multiple SNPs, which are often correlated and function jointly, and might be large in number. However, only a sparse subset of SNPs in a genetic construct is generally associated with the disease of interest. In this article, we propose the generalized higher criticism (GHC) to test for the association between an SNP set and a disease outcome. The higher criticism is a test traditionally used in high-dimensional signal detection settings when marginal test statistics are independent and the number of parameters is very large. However, these assumptions do not always hold in genetic association studies, due to linkage disequilibrium among SNPs and the finite number of SNPs in an SNP set in each genetic construct. The proposed GHC overcomes the limitations of the higher criticism by allowing for arbitrary correlation structures among the SNPs in an SNP-set, while performing accurate analytic p-value calculations for any finite number of SNPs in the SNP-set. We obtain the detection boundary of the GHC test. We compared empirically using simulations the power of the GHC method with existing SNP-set tests over a range of genetic regions with varied correlation structures and signal sparsity. We apply the proposed methods to analyze the CGEM breast cancer genome-wide association study. Supplementary materials for this article are available online.

Characterizing the clinical relevance of digital phenotyping data quality with applications to a cohort with schizophrenia

Digital phenotyping, or the moment-by-moment quantification of the individual-level human phenotype in situ using data from personal digital devices and smartphones, in particular, holds great potential for behavioral monitoring of patients. However, realizing the potential of digital phenotyping requires understanding of the smartphone as a scientific data collection tool. In this pilot study, we detail a procedure for estimating data quality for phone sensor samples and model the relationship between data quality and future symptom-related survey responses in a cohort with schizophrenia. We find that measures of empirical coverage of collected accelerometer and GPS data, as well as survey timing and survey completion metrics, are significantly associated with future survey scores for a variety of symptom domains. We also find evidence that specific measures of data quality are indicative of domain-specific future survey outcomes. These results suggest that for smartphone-based digital phenotyping, metadata is not independent of patient-reported survey scores, and is therefore potentially useful in predicting future clinical outcomes. This work raises important questions and considerations for future studies; we explore and discuss some of these implications.Digital phenotyping: assessing data quality in schizophreniaA pilot study shows that smartphone-collected data from patients with schizophrenia could be used to infer their mental-health status. Using smartphones as scientific data gathering tools holds great promise for understanding some of the behavioral features of psychiatric disorders and could provide an early indication of worsening symptoms. However, few studies have assessed the quality of the collected data, and thus the accuracy of clinical outcome prediction. Patrick Staples at the Harvard T. H. Chan School of Public Health in Boston, MA, and colleagues examined the relationship between data quality and future symptom-related survey responses in 16 patients with schizophrenia. They found that smartphone sensor data as well as phone-use metrics related to the completion of symptom-related surveys were significantly associated with survey results, highlighting the clinical relevance of this approach.

Relapse prediction in schizophrenia through digital phenotyping: a pilot study

Among individuals diagnosed, hospitalized, and treated for schizophrenia, up to 40% of those discharged may relapse within 1 year even with appropriate treatment. Passively collected smartphone behavioral data present a scalable and at present underutilized opportunity to monitor patients in order to identify possible warning signs of relapse. Seventeen patients with schizophrenia in active treatment at a state mental health clinic in Boston used the Beiwe app on their personal smartphone for up to 3 months. By testing for changes in mobility patterns and social behavior over time as measured through smartphone use, we were able to identify statistically significant anomalies in patient behavior in the days prior to relapse. We found that the rate of behavioral anomalies detected in the 2 weeks prior to relapse was 71% higher than the rate of anomalies during other time periods. Our findings show how passive smartphone data, data collected in the background during regular phone use without active input from the subjects, can provide an unprecedented and detailed view into patient behavior outside the clinic. Real-time detection of behavioral anomalies could signal the need for an intervention before an escalation of symptoms and relapse occur, therefore reducing patient suffering and reducing the cost of care.

A comparison of passive and active estimates of sleep in a cohort with schizophrenia

Sleep abnormalities are considered an important feature of schizophrenia, yet convenient and reliable sleep monitoring remains a challenge. Smartphones offer a novel solution to capture both self-reported and objective measures of sleep in schizophrenia. In this three-month observational study, 17 subjects with a diagnosis of schizophrenia currently in treatment downloaded Beiwe, a platform for digital phenotyping, on their personal Apple or Android smartphones. Subjects were given tri-weekly ecological momentary assessments (EMAs) on their own smartphones, and passive data including accelerometer, GPS, screen use, and anonymized call and text message logs was continuously collected. We compare the in-clinic assessment of sleep quality, assessed with the Pittsburgh Sleep Questionnaire Inventory (PSQI), to EMAs, as well as sleep estimates based on passively collected accelerometer data. EMAs and passive data classified 85% (11/13) of subjects as exhibiting high or low sleep quality compared to the in-clinic assessments among subjects who completed at least one in-person PSQI. Phone-based accelerometer data used to infer sleep duration was moderately correlated with subject self-assessment of sleep duration (r = 0.69, 95% CI 0.23–0.90). Active and passive phone data predicts concurrent PSQI scores for all subjects with mean average error of 0.75 and future PSQI scores with a mean average error of 1.9, with scores ranging from 0–14. These results suggest sleep monitoring via personal smartphones is feasible for subjects with schizophrenia in a scalable and affordable manner.Patient monitoring: Smartphones can track schizophrenia-related sleep abnormalitiesSmartphones may one-day offer accessible, clinically-useful insights into schizophrenia patients’ sleep quality. Despite the clinical relevance of sleep to disease severity, monitoring technologies still evade convenience and reliability. In search of a preferential method, a group of Harvard University researchers led by Patrick Staples investigated the validity of data collected via patients’ own mobile phones. The team, with a cohort of 17 schizophrenia patients, compared the quality of data produced by smartphone sensors and smartphone-delivered questionnaires to that of an in-clinic evaluation. The results significantly showed that smartphone monitoring could generate information that approached the accuracy of in-clinic assessments. The team noted some areas for improvement; however, this study provides convincing justifications for further research into this non-invasive, low-cost, scalable method to monitor the sleep quality of schizophrenic patients.Sleep abnormalities are considered an important feature of schizophrenia, yet convenient and reliable sleep monitoring remains a challenge. Smartphones offer a novel solution to capture both self-reported and objective measures of sleep in schizophrenia. In this three-month observational study, 17 subjects with a diagnosis of schizophrenia currently in treatment downloaded Beiwe, a platform for digital phenotyping, on their personal Apple or Android smartphones. Subjects were given tri-weekly ecological momentary assessments (EMAs) on their own smartphones, and passive data including accelerometer, GPS, screen use, and anonymized call and text message logs was continuously collected. We compare the in-clinic assessment of sleep quality, assessed with the Pittsburgh Sleep Questionnaire Inventory (PSQI), to EMAs, as well as sleep estimates based on passively collected accelerometer data. EMAs and passive data classified 85% (11/13) of subjects as exhibiting high or low sleep quality compared to the in-clinic assessments among subjects who completed at least one in-person PSQI. Phone-based accelerometer data used to infer sleep duration was moderately correlated with subject self-assessment of sleep duration (r = 0.69, 95% CI 0.23–0.90). Active and passive phone data predicts concurrent PSQI scores for all subjects with mean average error of 0.75 and future PSQI scores with a mean average error of 1.9, with scores ranging from 0–14. These results suggest sleep monitoring via personal smartphones is feasible for subjects with schizophrenia in a scalable and affordable manner.

Social and Spatial Clustering of People at Humanity’s Largest Gathering

Macroscopic behavior of scientific and societal systems results from the aggregation of microscopic behaviors of their constituent elements, but connecting the macroscopic with the microscopic in human behavior has traditionally been difficult. Manifestations of homophily, the notion that individuals tend to interact with others who resemble them, have been observed in many small and intermediate size settings. However, whether this behavior translates to truly macroscopic levels, and what its consequences may be, remains unknown. Here, we use call detail records (CDRs) to examine the population dynamics and manifestations of social and spatial homophily at a macroscopic level among the residents of 23 states of India at the Kumbh Mela, a 3-month-long Hindu festival. We estimate that the festival was attended by 61 million people, making it the largest gathering in the history of humanity. While we find strong overall evidence for both types of homophily for residents of different states, participants from low-representation states show considerably stronger propensity for both social and spatial homophily than those from high-representation states. These manifestations of homophily are amplified on crowded days, such as the peak day of the festival, which we estimate was attended by 25 million people. Our findings confirm that homophily, which here likely arises from social influence, permeates all scales of human behavior.

A crossroad for validating digital tools in schizophrenia and mental health

Schizophrenia remains one of the most devastating chronic illnesses, impacting nearly 1.5% of the global population and creating an economic burden of up to 1.65% gross domestic product. It is not surprising that digital tools for schizophrenia, often smartphone-based software in the form of apps, have received so much recent attention and enthusiasm. Digital phenotyping holds tremendous potential in elucidating the complex heterogeneity of what we call schizophrenia and would therefore advance research. On a clinical front, smartphone data may become increasingly valuable in monitoring course and treatment response given the fact that these patients frequently have difficulties in adherence with clinical visits, and are often poor historians. However, the power of this paradigm is currently fueled more by the increasing ubiquity of technology than breakthroughs in clinical science. The accessibility and affordability of digital care derives from increasing global ownership of smartphones: it is estimated that six billion smartphones will be in circulation worldwide by 2020. As devices and sensors become ever cheaper and more sophisticated, the ability to capture a plethora of relevant data and deliver a myriad of content via network connectivity will continue to fuel the potential of digital approaches in mental health. As validation and reproducibility lags behind enthusiasm and availability, the potential clinical impact of these digital tools is at a crossroads. How might this new approach advance clinical care? Affordable and accurate diagnostics from smartphones paired with ondemand or automatically deployed interventions enables unprecedented access to mental health services. Apps today are designed to perform a wide range of healthcare tasks ranging from telehealth to medication tracking. In addition, new platforms are currently being developed to measure novel behavioral and physiological markers using passive long-term smartphone data, enabling objective measurement without burden for patients and healthcare providers. Clinically relevant and passively collected smartphone data comprises a wide range of sensors, including accelerometer data to estimate activity, anonymized call/text log information to estimate sociability, and screen touch data to estimate cognition. Passive measurement might also be able to distinguish disease subtypes to help better classify psychotic illnesses, similar to recent research using genetic, physiological, and cognitive markers. The considerable potential impact of these tools is paralleled by substantial new challenges. Formidable analytical complexities accompany all data-driven approaches, including clinical inference using passive smartphone data. Missing data, the highdimensional and temporally dense nature of the collected data, habits of smartphone use, quality of user experience with the app, and the quality of the software implementation may all act as confounders to underlying clinical disease state. Estimates of clinical accuracy and efficacy of these devices remains broadly understudied. Daunting implementation challenges also accompany this smartphone-based work: simple questions such as which patients are comfortable with smartphone monitoring, how long should it be used for, how information should be shared with patients and psychiatrists all remain largely unknown. Ethical questions also remain with respect to appropriate storage, access, and usage protocols for this highly personal data. Ignoring these challenges and questions would be both a scientific mistake and also a missed opportunity for clinical care. Recall the humorous 2009 case report of the dead North Atlantic Salmon, who was asked to detect emotions in photos during a fMRI task, resulting in the “finding” of correlated neural activity— due to failing to control for multiple comparisons. Online analyses of digital phenotyping data performed on say, a daily basis, are faced with the same challenge of correcting for multiple comparisons. The more recent discovery of widespread statistical software issues in thousands of fMRI research protocols underscores how simple mistakes can be amplified with digital tools. The promise of fMRI is owed to its high-resolution detail, which is directly tied to data complexity and the peril of inappropriate statistical inference. Digital phenotyping data, which is in situ, multi-sensor, partially observed, and longitudinal, brings even more complexity to bear, and deserves proportionate circumspection. To meet these challenges and to hone this approach into clinically useful tools, the development of research platforms must be accompanied with empirical research on the properties of the data it collects, called metadata. A simple example of metadata is the time it may take you respond to a smartphone query, instead of the response to the query itself. Studying metadata is useful for two reasons. First, understanding the limitations and biases of our tools used to draw clinical inferences will improve their specificity and clarify where they can be beneficially used. Second, metadata itself might offer novel insights about patient behavior and especially cognition that is not available using traditional metrics or evaluations. Recent research from our group suggests that properties of smartphone data may be more complex than often portrayed in the popular press, which holds relevance for clinical use. We show a correlation in the outcomes between metadata such as accelerometer coverage, GPS coverage, and survey completion timings, and future responses to questions about mood, anxiety, and psychotic symptoms. This might be evidence that these measures might help predict disease progression. We also find that some of these measures differ by operating system (iOS vs. Android), potentially indicating confounding by operating system or other variables, such as socioeconomic status. In addition to understanding metadata, the development of digital phenotyping tools will benefit from other considerations. Data standards and extensive testing may take significant forethought and precious resources, but such care typically helps the success of deployment in the long run. Without standards in collecting, processing, and reporting for digital phenotyping data, the end result might be a continuation of the pilot studies we currently see, which are expensive and their results are often left unreplicated. This pales in comparison to the value of highthroughput, well-coordinated, multi-site research efforts seen in

Beyond smartphones and sensors: choosing appropriate statistical methods for the analysis of longitudinal data

Objectives As smartphones and sensors become more prominently used in mobile health, the methods used to analyze the resulting data must also be carefully considered. The advantages of smartphone-based studies, including large quantities of temporally dense longitudinally captured data, must be matched with the appropriate statistical methods in order draw valid conclusions. In this paper, we review and provide recommendations in 3 critical domains of analysis for these types of temporally dense longitudinal data and highlight how misleading results can arise from improper use of these methods. Target Audience Clinicians, biostatisticians, and data analysts who have digital phenotyping data or are interested in performing a digital phenotyping study or any other type of longitudinal study with frequent measurements taken over an extended period of time. Scope We cover the following topics: 1) statistical models using longitudinal repeated measures, 2) multiple comparisons of correlated tests, and 3) dimension reduction for correlated behavioral covariates. While these 3 classes of methods are frequently used in digital phenotyping data analysis, we demonstrate via actual clinical studies data that they may sometimes not perform as expected when applied to novel digital data.