BulkLMM: Real-time genome scans for multiple quantitative traits using linear mixed models

Genetic studies often collect data using high-throughput phenotyping. That has led to the need for fast genomewide scans for large number of traits using linear mixed models (LMMs). Computing the scans one by one on each trait is time consuming. We have developed new algorithms for performing genome scans on a large number of quantitative traits using LMMs, BulkLMM, that speeds up the computation by orders of magnitude compared to one trait at a time scans. On a mouse BXD Liver Proteome data with more than 35,000 traits and 7,000 markers, BulkLMM completed in a few seconds. We use vectorized, multi-threaded operations and regularization to improve optimization, and numerical approximations to speed up the computations. Our software implementation in the Julia programming language also provides permutation testing for LMMs and is available at https://github.com/senresearch/BulkLMM.jl.

Matrix Linear Models for connecting metabolite composition to individual characteristics

High-throughput metabolomics data provide a detailed molecular window into biological processes. We consider the problem of assessing how the association of metabolite levels with individual (sample) characteristics such as sex or treatment may depend on metabolite characteristics such as pathway. Typically this is one in a two-step process: In the first step we assess the association of each metabolite with individual characteristics. In the second step an enrichment analysis is performed by metabolite characteristics among significant associations. We combine the two steps using a bilinear model based on the matrix linear model (MLM) framework we have previously developed for high-throughput genetic screens. Our framework can estimate relationships in metabolites sharing known characteristics, whether categorical (such as type of lipid or pathway) or numerical (such as number of double bonds in triglycerides). We demonstrate how MLM offers flexibility and interpretability by applying our method to three metabolomic studies. We show that our approach can separate the contribution of the overlapping triglycerides characteristics, such as the number of double bonds and the number of carbon atoms. The proposed method have been implemented in the open-source Julia package, MatrixLM. Data analysis scripts with example data analyses are also available.

Effect of statin treatment on metabolites, lipids and prostanoids in patients with Statin Associated Muscle Symptoms (SAMS)

BACKGROUND

Between 5-10% of patients discontinue statin therapy due to statin-associated adverse reactions, primarily statin associated muscle symptoms (SAMS). The absence of a clear clinical phenotype or of biomarkers poses a challenge for diagnosis and management of SAMS. Similarly, our incomplete understanding of the pathogenesis of SAMS hinders the identification of treatments for SAMS. Metabolomics, the profiling of metabolites in biofluids, cells and tissues is an important tool for biomarker discovery and provides important insight into the origins of symptomatology. In order to better understand the pathophysiology of this common disorder and to identify biomarkers, we undertook comprehensive metabolomic and lipidomic profiling of plasma samples from patients with SAMS who were undergoing statin rechallenge as part of their clinical care.

METHODS AND FINDINGS

We report our findings in 67 patients, 28 with SAMS (cases) and 39 statin-tolerant controls. SAMS patients were studied during statin rechallenge and statin tolerant controls were studied while on statin. Plasma samples were analyzed using untargeted LC-MS metabolomics and lipidomics to detect differences between cases and controls. Differences in lipid species in plasma were observed between cases and controls. These included higher levels of linoleic acid containing phospholipids and lower ether lipids and sphingolipids. Reduced levels of acylcarnitines and altered amino acid profile (tryptophan, tyrosine, proline, arginine, and taurine) were observed in cases relative to controls. Pathway analysis identified significant increase of urea cycle metabolites and arginine and proline metabolites among cases along with downregulation of pathways mediating oxidation of branched chain fatty acids, carnitine synthesis, and transfer of acetyl groups into mitochondria.

CONCLUSIONS

The plasma metabolome of patients with SAMS exhibited reduced content of long chain fatty acids and increased levels of linoleic acid (18:2) in phospholipids, altered energy production pathways (β-oxidation, citric acid cycle and urea cycles) as well as reduced levels of carnitine, an essential mediator of mitochondrial energy production. Our findings support the hypothesis that alterations in pro-inflammatory lipids (arachidonic acid pathway) and impaired mitochondrial energy metabolism underlie the muscle symptoms of patients with statin associated muscle symptoms (SAMS).

Changes in self-reported and accelerometer-measured physical activity among pregnant TRICARE Beneficiaries

Background

Physical activity is recommended for all pregnant individuals and can prevent excessive gestational weight gain. However, physical activity has not been assessed among military personnel and other TRICARE beneficiaries, who experience unique military lifestyles. The current study assessed physical activity among pregnant TRICARE beneficiaries, both active duty and non-active duty, as measured by accelerometry and self-report data to examine potential predictors of physical activity engagement in the third trimester, and if self-report data was consistent with accelerometry data. We expected having a lower BMI, being active-duty, and having higher baseline physical activity engagement to be associated with higher physical activity at 32-weeks. We also hypothesized that accelerometry data would show lower physical activity levels than the self-reported measure.

Methods

Participants were 430 TRICARE adult beneficiaries (204 Active Duty; 226 non-Active Duty) in San Antonio, TX who were part of a randomized controlled parent study that implemented a stepped-care behavioral intervention. Participants were recruited if they were less than 12-weeks gestation and did not have health conditions precluding dietary or physical activity changes (e.g., uncontrolled cardiovascular conditions) or would contribute to weight changes. Participants completed self-report measures and wore an Actical Activity Monitor accelerometer on their wrist to collect physical activity data at baseline and 32-weeks gestation.

Results

Based on the accelerometer data, 99% of participants were meeting moderate physical activity guidelines recommending 150 min of moderate activity per week at baseline, and 96% were meeting this recommendation at 32-weeks. Based on self-report data, 88% of participants at baseline and 92% at 32-weeks met moderate physical activity recommendations. Linear regression and zero-inflated negative binomial models indicated that baseline physical activity engagement predicted moderate physical activity later in pregnancy above and beyond BMI and military status. Surprisingly, self-reported data, but not accelerometer data, showed that higher baseline activity was associated with decreased vigorous activity at 32-weeks gestation. Additionally, self-report and accelerometry data had small correlations at baseline, but not at 32-weeks.

Conclusions

Future intervention efforts may benefit from intervening with individuals with lower pre-pregnancy activity levels, as those who are active seem to continue this habit.

Trial Registration

The trial is registered on clinicaltrials.gov (NCT 03057808).

Speeding up eQTL scans in the BXD population using GPUs

The BXD family of mouse strains are an important reference population for systems biology and genetics that have been fully sequenced and deeply phenotyped. To facilitate interactive use of genotype-phenotype relations using many massive omics data sets for this and other segregating populations, we have developed new algorithms and code that enable near-real-time whole-genome quantitative trait locus (QTL) scans for up to one million traits. By using easily parallelizable operations including matrix multiplication, vectorized operations, and element-wise operations, our method is more than 700 times faster than a R/qtl linear model genome scan using 16 threads. We used parallelization of different CPU threads as well as GPUs. We found that the speed advantage of GPUs is dependent on problem size and shape (the number of cases, number of genotypes, and number of traits). Our approach is ideal for interactive web services, such as GeneNetwork.org that need to display results in real-time. Our implementation is available as the Julia language package LiteQTL at https://github.com/senresearch/LiteQTL.jl.

Assessing the Contribution of Self-Monitoring Through a Commercial Weight Loss App: Mediation and Predictive Modeling Study

Background

Electronic self-monitoring technology has the potential to provide unique insights into important behaviors for inducing weight loss.

Objective

The aim of this study is to investigate the effects of electronic self-monitoring behavior (using the commercial Lose It! app) and weight loss interventions (with differing amounts of counselor feedback and support) on 4- and 12-month weight loss.

Methods

In this secondary analysis of the Fit Blue study, we compared the results of two interventions of a randomized controlled trial. Counselor-initiated participants received consistent support from the interventionists, and self-paced participants received assistance upon request. The participants (N=191), who were active duty military personnel, were encouraged to self-monitor their diet and exercise with the Lose It! app or website. We examined the associations between intervention assignment and self-monitoring behaviors. We conducted a mediation analysis of the intervention assignment for weight loss through multiple mediators—app use (calculated from the first principal component [PC] of electronically collected variables), number of weigh-ins, and 4-month weight change. We used linear regression to predict weight loss at 4 and 12 months, and the accuracy was measured using cross-validation.

Results

On average, the counselor-initiated–treatment participants used the app more frequently than the self-paced–treatment participants. The first PC represented app use frequencies, the second represented calories recorded, and the third represented reported exercise frequency and exercise caloric expenditure. We found that 4-month weight loss was partially mediated through app use (ie, the first PC; 60.3%) and the number of weigh-ins (55.8%). However, the 12-month weight loss was almost fully mediated by 4-month weight loss (94.8%). Linear regression using app data from the first 8 weeks, the number of self–weigh-ins at 8 weeks, and baseline data explained approximately 30% of the variance in 4-month weight loss. App use frequency (first PC; P=.001), self-monitored caloric intake (second PC; P=.001), and the frequency of self-weighing at 8 weeks (P=.008) were important predictors of 4-month weight loss. Predictions for 12-month weight with the same variables produced an R² value of 5%; only the number of self–weigh-ins was a significant predictor of 12-month weight loss. The R² value using 4-month weight loss as a predictor was 31%. Self-reported exercise did not contribute to either model (4 months: P=.77; 12 months: P=.15).

Conclusions

We found that app use and daily reported caloric intake had a substantial impact on weight loss prediction at 4 months. Our analysis did not find evidence of an association between participant self-monitoring exercise information and weight loss. As 12-month weight loss was completely mediated by 4-month weight loss, intervention targets should focus on promoting early and frequent dietary intake self-monitoring and self-weighing to promote early weight loss, which leads to long-term success.

Trial Registration

ClinicalTrials.gov NCT02063178; https://clinicaltrials.gov/ct2/show/NCT02063178

Importance of Multiple Reinforcing Comments and Areas for Change in Optimizing Dietary and Exercise Self-Monitoring Feedback in Behavioral Weight Loss Programs: Factorial Design

Background

Individualized dietary and physical activity self-monitoring feedback is a core element of behavioral weight loss interventions and is associated with clinically significant weight loss. To our knowledge, no studies have evaluated individuals’ perspectives on the composition of feedback messages or the effect of feedback composition on the motivation to self-monitor.

Objective

This study aims to assess the perceptions of feedback emails as a function of the number of comments that reinforce healthy behavior and the number of areas for change (ie, behavioral changes that the individual might make to have an impact on weight) identified.

Methods

Emailed feedback followed a factorial design with 2 factors (ie, reinforcing comments and areas for change), each with 3 levels (ie, 1, 4, or 8 comments). A total of 250 adults with overweight or obesity who were interested in weight loss were recruited from the Qualtrics research panel. Participants read 9 emails presented in a random order. For each email, respondents answered 8 questions about the likelihood to self-monitor in the future, motivation for behavioral change, and perceptions of the counselor and the email. A mixed effects ordinal logistic model was used to compute conditional odds ratios and predictive margins (ie, average predicted probability) on a 5-point Likert response scale to investigate the optimal combination level of the 2 factors.

Results

Emails with more reinforcing comments or areas for change were better received, with small incremental benefits for 8 reinforcing comments or areas for change versus 4 reinforcing comments or areas for change. Interactions indicated that the best combination for 3 of 8 outcomes assessed (ie, motivation to make behavioral changes, counselor’s concern for their welfare, and the perception that the counselor likes them) was the email with 8 reinforcing comments and 4 areas for change. Emails with 4 reinforcing comments and 4 areas for change resulted in the highest average probability of individuals who reported being very likely to self-monitor in the future.

Conclusions

The study findings suggest how feedback might be optimized for efficacy. Future studies should explore whether the composition of feedback email affects actual self-monitoring performance.