Variable selection in omics data: A practical evaluation of small sample sizes

Survival prediction landscape: an in-depth systematic literature review on activities, methods, tools, diseases, and databases

Survival prediction integrates patient-specific molecular information and clinical signatures to forecast the anticipated time of an event, such as recurrence, death, or disease progression. Survival prediction proves valuable in guiding treatment decisions, optimizing resource allocation, and interventions of precision medicine. The wide range of diseases, the existence of various variants within the same disease, and the reliance on available data necessitate disease-specific computational survival predictors. The widespread adoption of artificial intelligence (AI) methods in crafting survival predictors has undoubtedly revolutionized this field. However, the ever-increasing demand for more sophisticated and effective prediction models necessitates the continued creation of innovative advancements. To catalyze these advancements, it is crucial to bring existing survival predictors knowledge and insights into a centralized platform. The paper in hand thoroughly examines 23 existing review studies and provides a concise overview of their scope and limitations. Focusing on a comprehensive set of 90 most recent survival predictors across 44 diverse diseases, it delves into insights of diverse types of methods that are used in the development of disease-specific predictors. This exhaustive analysis encompasses the utilized data modalities along with a detailed analysis of subsets of clinical features, feature engineering methods, and the specific statistical, machine or deep learning approaches that have been employed. It also provides insights about survival prediction data sources, open-source predictors, and survival prediction frameworks.

Individual Predictors of Response to A Behavioral Activation-Based Digital Smoking Cessation Intervention: A Machine Learning Approach

Background: Depression is prevalent among individuals who smoke cigarettes and increases risk for relapse. A previous clinical trial suggests that Goal2Quit, a behavioral activation-based smoking cessation mobile app, effectively increases smoking abstinence and reduces depressive symptoms. Objective: Secondary analyses were conducted on these trial data to identify predictors of success in depression-specific digitalized cessation interventions. Methods: Adult who smoked cigarettes (age = 38.4 ± 10.3, 53% women) were randomized to either use Goal2Quit for 12 weeks (N = 103), paired with a 2-week sample of nicotine replacement therapy (patch and lozenge) or to a Treatment-As-Usual (TAU) control (N = 47). The least absolute shrinkage and selection operator was utilized to identify a subset of baseline variables predicting either smoking or depression intervention outcomes. The retained predictors were then fitted via linear regression models to determine relations to each intervention outcome. Results: Relative to TAU, only individuals who spent significant time using Goal2Quit (56 ± 46 min) were more likely to reduce cigarette use by at least 50% after 12 weeks, whereas those who spent minimal time using Goal2Quit (10 ± 2 min) did not exhibit significant changes. An interaction between educational attainment and treatment group revealed that, as compared to TAU, only app users with an educational degree beyond high school exhibited significant reductions in depression. Conclusions: The findings highlight the importance of tailoring depression-specific digital cessation interventions to individuals' unique engagement needs and educational level. This study provides a potential methodological template for future research aimed at personalizing technology-based treatments for cigarette users with depressive symptoms.

Role of oral health in heart and vascular health: A population-based study

BACKGROUND AND AIM

Conditions such as hypertension, cardiovascular diseases, and hypercholesterolemia, are a major public health challenge. This study investigates the influence of oral health indicators, including gum bleeding, active dental caries, tooth mobility, and tooth loss, on their prevalence in Hungary, considering socioeconomic, demographic, and lifestyle factors.

MATERIALS AND METHODS

Data from the 2019 Hungarian European Health Interview Survey with 5,603 participants informed this analysis. Data were accessed from the records maintained by the Department of Health Informatics at the University of Debrecen between September and November 2023. Variable selection employed elastic net regularization and k-fold cross-validation, leading to high-performing predictors for weighted multiple logistic regression models. Sensitivity analysis confirmed the findings' validity.

RESULTS

Significant links were found between poor oral health and chronic cardiac conditions. Multiple teeth extractions increased hypertension risk (OR = 1.67, 95% CI: [1.01-2.77]); dental prosthetics had an OR of 1.45 [1.20-1.75]. Gum bleeding was associated with higher cardiovascular disease (OR = 1.69 [1.30-2.21]) and hypercholesterolemia risks (OR = 1.40 [1.09-1.81]).

CONCLUSIONS

Oral health improvement may reduce the risk of cardiac conditions. This underscores oral health's role in multidisciplinary disease management.

Identifying and overcoming COVID-19 vaccination impediments using Bayesian data mining techniques

The COVID-19 pandemic has profoundly reshaped human life. The development of COVID-19 vaccines has offered a semblance of normalcy. However, obstacles to vaccination have led to substantial loss of life and economic burdens. In this study, we analyze data from a prominent health insurance provider in the United States to uncover the underlying reasons behind the inability, refusal, or hesitancy to receive vaccinations. Our research proposes a methodology for pinpointing affected population groups and suggests strategies to mitigate vaccination barriers and hesitations. Furthermore, we estimate potential cost savings resulting from the implementation of these strategies. To achieve our objectives, we employed Bayesian data mining methods to streamline data dimensions and identify significant variables (features) influencing vaccination decisions. Comparative analysis reveals that the Bayesian method outperforms cutting-edge alternatives, demonstrating superior performance.

Text-mining-based feature selection for anticancer drug response prediction

Motivation

Predicting anticancer treatment response from baseline genomic data is a critical obstacle in personalized medicine. Machine learning methods are commonly used for predicting drug response from gene expression data. In the process of constructing these machine learning models, one of the most significant challenges is identifying appropriate features among a massive number of genes.

Results

In this study, we utilize features (genes) extracted using the text-mining of scientific literatures. Using two independent cancer pharmacogenomic datasets, we demonstrate that text-mining-based features outperform traditional feature selection techniques in machine learning tasks. In addition, our analysis reveals that text-mining feature-based machine learning models trained on in vitro data also perform well when predicting the response of in vivo cancer models. Our results demonstrate that text-mining-based feature selection is an easy to implement approach that is suitable for building machine learning models for anticancer drug response prediction.

Availability and implementation

https://github.com/merlab/text_features.

When stroke survivors' self-ratings are inconsistent with the ratings of others: a cohort study examining biopsychosocial factors associated with impaired self-awareness of functional abilities

Background Stroke survivors' self-ratings of functional abilities are often inconsistent with ratings assigned by others (e.g. clinicians), a phenomenon referred to as 'impaired self-awareness' (ISA). There is limited knowledge of the biopsychosocial contributors and consequences of post-stroke ISA measured across the rehabilitation journey. This multi-site cohort study explored biopsychosocial correlates of ISA during subacute rehabilitation (inpatient) and at 4 months post-discharge (community-dwelling). Methods Forty-five subacute stroke survivors participated (Age M (s.d.) =  71.5 (15.6), 56% female), and 38 were successfully followed-up. Self-assessments were compared to those of an independent rater (occupational therapist, close other) to calculate ISA at both time points. Survivors and raters completed additional cognitive, psychological and functional measures. Results Multivariate regression (multiple outcomes) identified associations between ISA during inpatient admission and poorer outcomes at follow-up, including poorer functional cognition, participation restriction, caregiver burden, and close other depression and anxiety. Regression models applied cross-sectionally, including one intended for correlated predictors, indicated associations between ISA during inpatient admission and younger age, male sex, poorer functional cognition, poorer rehabilitation engagement and less frequent use of non-productive coping (adjusted R 2  = 0.60). ISA at community follow-up was associated with poorer functional cognition and close other anxiety (adjusted R 2  = 0.66). Conclusions Associations between ISA and poorer outcomes across the rehabilitation journey highlight the clinical importance of ISA and the value of assessment and management approaches that consider the potential influence of numerous biological and psychosocial factors on ISA. Future studies should use larger sample sizes to confirm these results and determine the causal mechanisms of these relationships.

Metabolic consequences of various fruit-based diets in a generalist insect species

Most phytophagous insect species exhibit a limited diet breadth and specialize on a few or a single host plant. In contrast, some species display a remarkably large diet breadth, with host plants spanning several families and many species. It is unclear, however, whether this phylogenetic generalism is supported by a generic metabolic use of common host chemical compounds ('metabolic generalism') or alternatively by distinct uses of diet-specific compounds ('multi-host metabolic specialism')? Here, we simultaneously investigated the metabolomes of fruit diets and of individuals of a generalist phytophagous species, Drosophila suzukii, that developed on them. The direct comparison of metabolomes of diets and consumers enabled us to disentangle the metabolic fate of common and rarer dietary compounds. We showed that the consumption of biochemically dissimilar diets resulted in a canalized, generic response from generalist individuals, consistent with the metabolic generalism hypothesis. We also showed that many diet-specific metabolites, such as those related to the particular color, odor, or taste of diets, were not metabolized, and rather accumulated in consumer individuals, even when probably detrimental to fitness. As a result, while individuals were mostly similar across diets, the detection of their particular diet was straightforward. Our study thus supports the view that dietary generalism may emerge from a passive, opportunistic use of various resources, contrary to more widespread views of an active role of adaptation in this process. Such a passive stance towards dietary chemicals, probably costly in the short term, might favor the later evolution of new diet specializations.

Persistent somatic symptoms are key to individual illness perception at one year after COVID-19 in a cross-sectional analysis of a prospective cohort study

OBJECTIVE

Subjective illness perception (IP) can differ from physician's clinical assessment results. Herein, we explored patient's IP during coronavirus disease 2019 (COVID-19) recovery.

METHODS

Participants of the prospective observation CovILD study (ClinicalTrials.gov: NCT04416100) with persistent somatic symptoms or cardiopulmonary findings one year after COVID-19 were analyzed (n = 74). Explanatory variables included demographic and comorbidity, COVID-19 course and one-year follow-up data of persistent somatic symptoms, physical performance, lung function testing, chest computed tomography and trans-thoracic echocardiography. Factors affecting IP (Brief Illness Perception Questionnaire) one year after COVID-19 were identified by regularized modeling and unsupervised clustering.

RESULTS

In modeling, 33% of overall IP variance (R2) was attributed to fatigue intensity, reduced physical performance and persistent somatic symptom count. Overall IP was largely independent of lung and heart findings revealed by imaging and function testing. In clustering, persistent somatic symptom count (Kruskal-Wallis test: η2 = 0.31, p < .001), fatigue (η2 = 0.34, p < .001), diminished physical performance (χ2 test, Cramer V effect size statistic: V = 0.51, p < .001), dyspnea (V = 0.37, p = .006), hair loss (V = 0.57, p < .001) and sleep problems (V = 0.36, p = .008) were strongly associated with the concern, emotional representation, complaints, disease timeline and consequences IP dimensions.

CONCLUSION

Persistent somatic symptoms rather than abnormalities in cardiopulmonary testing influence IP one year after COVID-19. Modifying IP represents a promising innovative approach to treatment of post-COVID-19 condition. Besides COVID-19 severity, individual IP should guide rehabilitation and psychological therapy decisions.

Transcriptomes of the tumor-adjacent normal tissues are more informative than tumors in predicting recurrence in colorectal cancer patients

BACKGROUND

Previous investigations of transcriptomic signatures of cancer patient survival and post-therapy relapse have focused on tumor tissue. In contrast, here we show that in colorectal cancer (CRC) transcriptomes derived from normal tissues adjacent to tumors (NATs) are better predictors of relapse.

RESULTS

Using the transcriptomes of paired tumor and NAT specimens from 80 Korean CRC patients retrospectively determined to be in recurrence or nonrecurrence states, we found that, when comparing recurrent with nonrecurrent samples, NATs exhibit a greater number of differentially expressed genes (DEGs) than tumors. Training two prognostic elastic net-based machine learning models-NAT-based and tumor-based in our Samsung Medical Center (SMC) cohort, we found that NAT-based model performed better in predicting the survival when the model was applied to the tumor-derived transcriptomes of an independent cohort of 450 COAD patients in TCGA. Furthermore, compositions of tumor-infiltrating immune cells in NATs were found to have better prognostic capability than in tumors. We also confirmed through Cox regression analysis that in both SMC-CRC as well as in TCGA-COAD cohorts, a greater proportion of genes exhibited significant hazard ratio when NAT-derived transcriptome was used compared to when tumor-derived transcriptome was used.

CONCLUSIONS

Taken together, our results strongly suggest that NAT-derived transcriptomes and immune cell composition of CRC are better predictors of patient survival and tumor recurrence than the primary tumor.

Proteomics biomarker discovery for individualized prevention of familial pancreatic cancer using statistical learning

BACKGROUND

The low five-year survival rate of pancreatic ductal adenocarcinoma (PDAC) and the low diagnostic rate of early-stage PDAC via imaging highlight the need to discover novel biomarkers and improve the current screening procedures for early diagnosis. Familial pancreatic cancer (FPC) describes the cases of PDAC that are present in two or more individuals within a circle of first-degree relatives. Using innovative high-throughput proteomics, we were able to quantify the protein profiles of individuals at risk from FPC families in different potential pre-cancer stages. However, the high-dimensional proteomics data structure challenges the use of traditional statistical analysis tools. Hence, we applied advanced statistical learning methods to enhance the analysis and improve the results' interpretability.

METHODS

We applied model-based gradient boosting and adaptive lasso to deal with the small, unbalanced study design via simultaneous variable selection and model fitting. In addition, we used stability selection to identify a stable subset of selected biomarkers and, as a result, obtain even more interpretable results. In each step, we compared the performance of the different analytical pipelines and validated our approaches via simulation scenarios.

RESULTS

In the simulation study, model-based gradient boosting showed a more accurate prediction performance in the small, unbalanced, and high-dimensional datasets than adaptive lasso and could identify more relevant variables. Furthermore, using model-based gradient boosting, we discovered a subset of promising serum biomarkers that may potentially improve the current screening procedure of FPC.

CONCLUSION

Advanced statistical learning methods helped us overcome the shortcomings of an unbalanced study design in a valuable clinical dataset. The discovered serum biomarkers provide us with a clear direction for further investigations and more precise clinical hypotheses regarding the development of FPC and optimal strategies for its early detection.

The Application of Spearman Partial Correlation for Screening Predictors of Weight Loss in a Multiomics Dataset

Obesity has reached epidemic proportions in the United States, but little is known about the mechanisms of weight gain and weight loss. Integration of omics data is becoming a popular tool to increase understanding in such complex phenotypes. Biomarkers come in abundance, but small sample size remains a serious limitation in clinical trials. In the present study, we developed a strategy to screen predictors from a multiomics, high-dimensional, and longitudinal dataset from a small cohort of 10 women with obesity who were provided an identical very-low calorie diet. Our proposal explores the combinatorial space of potential predictors from transcriptomics, microbiome, metabolome, fecal bile acids, and clinical data with the application of the first-order Spearman partial correlation coefficient. Two statistics are proposed for screening predictors, the partial association score, and the persistent significance. We applied our strategy to predict rates of weight loss in our sample of participants in a hospital metabolic facility. Our method reduced an initial set of 42,000 biomarker candidates to 61 robust predictors. The results show baseline fecal bile acids and regulation in RT-polymerase chain reaction as the most predictive data sources in forecasting the rate of weight-loss. In summary, the present study proposes a strategy based on nonparametric statistics for ranking and screening predictors of weight loss from a multiomics study. The proposed biomarker screening strategy warrants further translational clinical investigation in obesity and other complex clinical phenotypes.

Obesity-Associated Differentially Methylated Regions in Colon Cancer

Obesity with adiposity is a common disorder in modern days, influenced by environmental factors such as eating and lifestyle habits and affecting the epigenetics of adipose-based gene regulations and metabolic pathways in colorectal cancer (CRC). We compared epigenetic changes of differentially methylated regions (DMR) of genes in colon tissues of 225 colon cancer cases (154 non-obese and 71 obese) and 15 healthy non-obese controls by accessing The Cancer Genome Atlas (TCGA) data. We applied machine-learning-based analytics including generalized regression (GR) as a confirmatory validation model to identify the factors that could contribute to DMRs impacting colon cancer to enhance prediction accuracy. We found that age was a significant predictor in obese cancer patients, both alone (p = 0.003) and interacting with hypomethylated DMRs of ZBTB46, a tumor suppressor gene (p = 0.008). DMRs of three additional genes: HIST1H3I (p = 0.001), an oncogene with a hypomethylated DMR in the promoter region; SRGAP2C (p = 0.006), a tumor suppressor gene with a hypermethylated DMR in the promoter region; and NFATC4 (p = 0.006), an adipocyte differentiating oncogene with a hypermethylated DMR in an intron region, are also significant predictors of cancer in obese patients, independent of age. The genes affected by these DMR could be potential novel biomarkers of colon cancer in obese patients for cancer prevention and progression.

Pan-Cancer Analysis Reveals Alternative Splicing Characteristics Associated With Immune-Related Adverse Events Elicited by Checkpoint Immunotherapy

Immune-related adverse events (irAEs) can impair the effectiveness and safety of immune checkpoint inhibitors (ICIs) and restrict the clinical applications of ICIs in oncology. The predictive biomarkers of irAE are urgently required for early diagnosis and subsequent management. The exact mechanism underlying irAEs remains to be fully elucidated, and the availability of predictive biomarkers is limited. Herein, we performed data mining by combining pharmacovigilance data and pan-cancer transcriptomic information to illustrate the relationships between alternative splicing characteristics and irAE risk of ICIs. Four distinct classes of splicing characteristics considered were associated with splicing factors, neoantigens, splicing isoforms, and splicing levels. Correlation analysis confirmed that expression levels of splicing factors were predictive of irAE risk. Adding DHX16 expression to the bivariate PD-L1 protein expression-fPD1 model markedly enhanced the prediction for irAE. Furthermore, we identified 668 and 1,131 potential predictors based on the correlation of the incidence of irAEs with splicing frequency and isoform expression, respectively. The functional analysis revealed that alternative splicing might contribute to irAE pathogenesis via coordinating innate and adaptive immunity. Remarkably, autoimmune-related genes and autoantigens were preferentially over-represented in these predictors for irAE, suggesting a close link between autoimmunity and irAE occurrence. In addition, we established a trivariate model composed of CDC42EP3-206, TMEM138-211, and IRX3-202, that could better predict the risk of irAE across various cancer types, indicating a potential application as promising biomarkers for irAE. Our study not only highlights the clinical relevance of alternative splicing for irAE development during checkpoint immunotherapy but also sheds new light on the mechanisms underlying irAEs.

Fatty acid metabolism and colon cancer protection by dietary methyl donor restriction

INTRODUCTION

A methyl donor depleted (MDD) diet dramatically suppresses intestinal tumor development in Apc-mutant mice, but the mechanism of this prevention is not entirely clear.

OBJECTIVES

We sought to gain insight into the mechanisms of cancer suppression by the MDD diet and to identify biomarkers of cancer risk reduction.

METHODS

A plasma metabolomic analysis was performed on Apc^Δ14/+ mice maintained on either a methyl donor sufficient (MDS) diet or the protective MDD diet. A group of MDS animals was also pair-fed with the MDD mice to normalize caloric intake, and another group was shifted from an MDD to MDS diet to determine the durability of the metabolic changes.

RESULTS

In addition to the anticipated changes in folate one-carbon metabolites, plasma metabolites related to fatty acid metabolism were generally decreased by the MDD diet, including carnitine, acylcarnitines, and fatty acids. Some fatty acid selectivity was observed; the levels of cancer-promoting arachidonic acid and 2-hydroxyglutarate were decreased by the MDD diet, whereas eicosapentaenoic acid (EPA) levels were increased. Machine-learning elastic net analysis revealed a positive association between the fatty acid-related compounds azelate and 7-hydroxycholesterol and tumor development, and a negative correlation with succinate and β-sitosterol.

CONCLUSION

Methyl donor restriction causes dramatic changes in systemic fatty acid metabolism. Regulating fatty acid metabolism through methyl donor restriction favorably effects fatty acid profiles to achieve cancer protection.

Enhancing data pipelines for forecasting student performance: integrating feature selection with cross-validation

Educators seek to harness knowledge from educational corpora to improve student performance outcomes. Although prior studies have compared the efficacy of data mining methods (DMMs) in pipelines for forecasting student success, less work has focused on identifying a set of relevant features prior to model development and quantifying the stability of feature selection techniques. Pinpointing a subset of pertinent features can (1) reduce the number of variables that need to be managed by stakeholders, (2) make "black-box" algorithms more interpretable, and (3) provide greater guidance for faculty to implement targeted interventions. To that end, we introduce a methodology integrating feature selection with cross-validation and rank each feature on subsets of the training corpus. This modified pipeline was applied to forecast the performance of 3225 students in a baccalaureate science course using a set of 57 features, four DMMs, and four filter feature selection techniques. Correlation Attribute Evaluation (CAE) and Fisher's Scoring Algorithm (FSA) achieved significantly higher Area Under the Curve (AUC) values for logistic regression (LR) and elastic net regression (GLMNET), compared to when this pipeline step was omitted. Relief Attribute Evaluation (RAE) was highly unstable and produced models with the poorest prediction performance. Borda's method identified grade point average, number of credits taken, and performance on concept inventory assessments as the primary factors impacting predictions of student performance. We discuss the benefits of this approach when developing data pipelines for predictive modeling in undergraduate settings that are more interpretable and actionable for faculty and stakeholders.

SUPPLEMENTARY INFORMATION

The online version contains supplementary material available at 10.1186/s41239-021-00279-6.

Plasma Amino Acids and Risk of Impaired Lower-Extremity Function and Role of Dietary Intake: A Nested Case-Control Study in Older Adults

INTRODUCTION

Amino acids are key elements in the regulation of the aging process which entails a progressive loss of muscle mass. The health effects of plasma amino acids can be influenced by dietary intake. This study assessed the prospective association between amino acid species and impaired lower-extremity function (ILEF) in older adults, exploring the role of diet on this association.

METHODS

This is a case-control design comprising 43 incident cases of ILEF and 85 age- and sex-matched controls. Plasma concentrations of 20 amino acid species were measured at baseline using liquid chromatography-tandem mass spectrometry, and incident cases of ILEF were measured after 2 years by means of the Short Physical Performance Battery. Conditional logistic regression models were used to assess longitudinal relationships.

RESULTS

After adjusting for potential confounders, higher levels of tryptophan were associated with a decreased 2-year risk of ILEF (OR per 1-SD increase = 0.64, 95% CI = [0.42, 0.97]), while glutamine and total essential amino acids were linked to higher ILEF risk (OR = 1.57, 95% CI = [1.01, 2.45]; OR = 1.89, 95% CI = [1.18, 3.03], respectively). Those with a lower adherence to a Mediterranean diet, a higher BMI, a higher consumption of red meat, and a lower consumption of nuts and legumes had an increased risk of ILEF associated with higher levels of essential amino acids.

DISCUSSION/CONCLUSION

Some amino acid species could serve as risk markers for physical function decline in older adults, and healthy diet might attenuate the excess risk of ILEF linked to essential amino acids.

Age-dependent increase in α-tocopherol and phytosterols in maize leaves exposed to elevated ozone pollution

Tropospheric ozone is a major air pollutant that significantly damages crop production. Crop metabolic responses to rising chronic ozone stress have not been well studied in the field, especially in C4 crops. In this study, we investigated the metabolomic profile of leaves from two diverse maize (Zea mays) inbred lines and the hybrid cross during exposure to season-long elevated ozone (~100 nl L-1) in the field using free air concentration enrichment (FACE) to identify key biochemical responses of maize to elevated ozone. Senescence, measured by loss of chlorophyll content, was accelerated in the hybrid line, B73 × Mo17, but not in either inbred line (B73 or Mo17). Untargeted metabolomic profiling further revealed that inbred and hybrid lines of maize differed in metabolic responses to ozone. A significant difference in the metabolite profile of hybrid leaves exposed to elevated ozone occurred as leaves aged, but no age-dependent difference in leaf metabolite profiles between ozone conditions was measured in the inbred lines. Phytosterols and α-tocopherol levels increased in B73 × Mo17 leaves as they aged, and to a significantly greater degree in elevated ozone stress. These metabolites are involved in membrane stabilization and chloroplast reactive oxygen species (ROS) quenching. The hybrid line also showed significant yield loss at elevated ozone, which the inbred lines did not. This suggests that the hybrid maize line was more sensitive to ozone exposure than the inbred lines, and up-regulated metabolic pathways to stabilize membranes and quench ROS in response to chronic ozone stress.

Can't decide how much to EAT? Effort variability for reward is associated with cognitive restraint

Food intake is inherently variable and often characterized by episodical restraint or overeating (uncontrolled eating). Such heightened variability in intake has been associated with higher variability in the brain response to food reward, but it is an open issue whether comparable associations with elevated variability in reward seeking exist. Here, we assessed whether restraint and uncontrolled eating as markers of trait-like variability in eating are associated with higher intra-individual variability in reward seeking as captured by a cost-benefit paradigm. To test this hypothesis, 81 healthy, overnight-fasting participants (M_BMI = 23.0 kg/m² ± 3.0) completed an effort allocation task (EAT) twice. In the EAT, participants had to exert physical effort to earn monetary and food rewards and indicated levels of wanting through visual analog scales (VAS). As predicted, we found that greater trial-by-trial effort variability was associated with lower scores on cognitive restraint, r_p(78) = -0.28, p = .011 (controlled for average effort). In line with previous findings, higher wanting variability was associated with higher BMI, r_p(78) = 0.25, p = .026 (controlled for average effort). Collectively, our results support the idea that higher variability in reward seeking is a potential risk factor for eating beyond homeostatic need. Since associations with variability measures of reward exceeded associations with average reward seeking, our findings may indicate that variability in the representation of the reward value could be a crucial aspect driving fluctuations in food intake.

Multiplexed quantitative proteomics provides mechanistic cues for malaria severity and complexity

Management of severe malaria remains a critical global challenge. In this study, using a multiplexed quantitative proteomics pipeline we systematically investigated the plasma proteome alterations in non-severe and severe malaria patients. We identified a few parasite proteins in severe malaria patients, which could be promising from a diagnostic perspective. Further, from host proteome analysis we observed substantial modulations in many crucial physiological pathways, including lipid metabolism, cytokine signaling, complement, and coagulation cascades in severe malaria. We propose that severe manifestations of malaria are possibly underpinned by modulations of the host physiology and defense machinery, which is evidently reflected in the plasma proteome alterations. Importantly, we identified multiple blood markers that can effectively define different complications of severe falciparum malaria, including cerebral syndromes and severe anemia. The ability of our identified blood markers to distinguish different severe complications of malaria may aid in developing new clinical tests for monitoring malaria severity.

Multi-omics prediction of immune-related adverse events during checkpoint immunotherapy

Immune-related adverse events (irAEs), caused by anti-PD-1/PD-L1 antibodies, can lead to fulminant and even fatal consequences and thus require early detection and aggressive management. However, a comprehensive approach to identify biomarkers of irAE is lacking. Here, we utilize a strategy that combines pharmacovigilance data and omics data, and evaluate associations between multi-omics factors and irAE reporting odds ratio across different cancer types. We identify a bivariate regression model of LCP1 and ADPGK that can accurately predict irAE. We further validate LCP1 and ADPGK as biomarkers in an independent patient-level cohort. Our approach provides a method for identifying potential biomarkers of irAE in cancer immunotherapy using both pharmacovigilance data and multi-omics data.

Immunotherapy, the reactivation of the immune system to recognize cancer cells, can be accompanied by severe adverse effects. Here, the authors use pharmacovigilance and genomic data to be able to predict which patients might be susceptible to such severe events.

Solid phase microextraction chemical biopsy tool for monitoring of doxorubicin residue during in vivo lung chemo-perfusion

Development of a novel in vivo lung perfusion (IVLP) procedure allows localized delivery of high-dose doxorubicin (DOX) for targeting residual micrometastatic disease in the lungs. However, DOX delivery via IVLP requires careful monitoring of drug level to ensure tissue concentrations of this agent remain in the therapeutic window. A small dimension nitinol wire coated with a sorbent of biocompatible morphology (Bio-SPME) has been clinically evaluated for in vivo lung tissue extraction and determination of DOX and its key metabolites. The in vivo Bio-SPME-IVLP experiments were performed on pig model over various (150 and 225 mg/m²) drug doses, and during human clinical trial. Two patients with metastatic osteosarcoma were treated with a single 5 and 7 μg/mL (respectively) dose of DOX during a 3-h IVLP. In both pig and human cases, DOX tissue levels presented similar trends during IVLP. Human lung tissue concentrations of drug ranged between 15 and 293 μg/g over the course of the IVLP procedure. In addition to DOX levels, Bio-SPME followed by liquid chromatography-mass spectrometry analysis generated 64 metabolic features during endogenous metabolite screening, providing information about lung status during drug administration. Real-time monitoring of DOX levels in the lungs can be performed effectively throughout the IVLP procedure by in vivo Bio-SPME chemical biopsy approach. Bio-SPME also extracted various endogenous molecules, thus providing a real-time snapshot of the physiology of the cells, which might assist in the tailoring of personalized treatment strategy.

Lipidomic Profiling of the Epidermis in a Mouse Model of Dermatitis Reveals Sexual Dimorphism and Changes in Lipid Composition before the Onset of Clinical Disease

Atopic dermatitis (AD) is a multifactorial disease associated with alterations in lipid composition and organization in the epidermis. Multiple variants of AD exist with different outcomes in response to therapies. The evaluation of disease progression and response to treatment are observational assessments with poor inter-observer agreement highlighting the need for molecular markers. SHARPIN-deficient mice (Sharpin^cpdm) spontaneously develop chronic proliferative dermatitis with features similar to AD in humans. To study the changes in the epidermal lipid-content during disease progression, we tested 72 epidermis samples from three groups (5-, 7-, and 10-weeks old) of cpdm mice and their WT littermates. An agnostic mass-spectrometry strategy for biomarker discovery termed multiple-reaction monitoring (MRM)-profiling was used to detect and monitor 1,030 lipid ions present in the epidermis samples. In order to select the most relevant ions, we utilized a two-tiered filter/wrapper feature-selection strategy. Lipid categories were compressed, and an elastic-net classifier was used to rank and identify the most predictive lipid categories for sex, phenotype, and disease stages of cpdm mice. The model accurately classified the samples based on phospholipids, cholesteryl esters, acylcarnitines, and sphingolipids, demonstrating that disease progression cannot be defined by one single lipid or lipid category.

Multi-Omic Regulation of the PAM50 Gene Signature in Breast Cancer Molecular Subtypes

Breast cancer is a disease that exhibits heterogeneity that goes from the genomic to the clinical levels. This heterogeneity is thought to be captured (at least partially) by the so-called breast cancer molecular subtypes. These molecular subtypes were initially defined based on the unsupervised clustering of gene expression and its correlate with histological, morphological, phenotypic and clinical features already known. Later, a 50-gene signature, PAM50, was defined in order to identify the biological subtype of a given sample within the clinical setting. The PAM50 signature was obtained by the use of unsupervised statistical methods, and therefore no limitation was set on the biological relevance (or lack of) of the selected genes beyond its predictive capacity. An open question that remains is what are the regulatory elements that drive the various expression behaviors of this set of genes in the different molecular subtypes. This question becomes more relevant as the measurement of more biological layers of regulation becomes accessible. In this work, we analyzed the gene expression regulation of the 50 genes in the PAM50 signature, in terms of (a) gene co-expression, (b) transcription factors, (c) micro-RNAs, and (d) methylation. Using data from the Cancer Genome Atlas (TCGA) for the Luminal A and B, Basal, and HER2-enriched molecular subtypes as well as normal tumor adjacent tissue, we identified predictors for gene expression through the use of an elastic net model. We compare and contrast the sets of identified regulators for the gene signature in each molecular subtype, and systematically compare them to current literature. We also identified a unique set of predictors for the expression of genes in the PAM50 signature associated with each of the molecular subtypes. Most selected predictors are exclusive for a PAM50 gene and predictors are not shared across subtypes. There are only 13 coding transcripts and 2 miRNAs selected for the four subtypes. MiR-21 and miR-10b connect almost all the PAM50 genes in all the subtypes and normal tissue, but do it in an exclusive manner, suggesting a cancer switch from miR-10b coordination in normal tissue to miR-21. The PAM50 gene sets of selected predictors that enrich for a function across subtypes, support that different regulatory molecular mechanisms are taking place. With this study we aim to a wider understanding of the regulatory mechanisms that differentiate the expression of the PAM50 signature, which in turn could perhaps help understand the molecular basis of the differences between the molecular subtypes.

MLGaze: Machine Learning-Based Analysis of Gaze Error Patterns in Consumer Eye Tracking Systems

Analyzing the gaze accuracy characteristics of an eye tracker is a critical task as its gaze data is frequently affected by non-ideal operating conditions in various consumer eye tracking applications. In previous research on pattern analysis of gaze data, efforts were made to model human visual behaviors and cognitive processes. What remains relatively unexplored are questions related to identifying gaze error sources as well as quantifying and modeling their impacts on the data quality of eye trackers. In this study, gaze error patterns produced by a commercial eye tracking device were studied with the help of machine learning algorithms, such as classifiers and regression models. Gaze data were collected from a group of participants under multiple conditions that commonly affect eye trackers operating on desktop and handheld platforms. These conditions (referred here as error sources) include user distance, head pose, and eye-tracker pose variations, and the collected gaze data were used to train the classifier and regression models. It was seen that while the impact of the different error sources on gaze data characteristics were nearly impossible to distinguish by visual inspection or from data statistics, machine learning models were successful in identifying the impact of the different error sources and predicting the variability in gaze error levels due to these conditions. The objective of this study was to investigate the efficacy of machine learning methods towards the detection and prediction of gaze error patterns, which would enable an in-depth understanding of the data quality and reliability of eye trackers under unconstrained operating conditions. Coding resources for all the machine learning methods adopted in this study were included in an open repository named MLGaze to allow researchers to replicate the principles presented here using data from their own eye trackers.

Transfer of motor skill between virtual reality viewed using a head-mounted display and conventional screen environments

BACKGROUND

Virtual reality viewed using a head-mounted display (HMD-VR) has the potential to be a useful tool for motor learning and rehabilitation. However, when developing tools for these purposes, it is important to design applications that will effectively transfer to the real world. Therefore, it is essential to understand whether motor skills transfer between HMD-VR and conventional screen-based environments and what factors predict transfer.

METHODS

We randomized 70 healthy participants into two groups. Both groups trained on a well-established measure of motor skill acquisition, the Sequential Visual Isometric Pinch Task (SVIPT), either in HMD-VR or in a conventional environment (i.e., computer screen). We then tested whether the motor skills transferred from HMD-VR to the computer screen, and vice versa. After the completion of the experiment, participants responded to questions relating to their presence in their respective training environment, age, gender, video game use, and previous HMD-VR experience. Using multivariate and univariate linear regression, we then examined whether any personal factors from the questionnaires predicted individual differences in motor skill transfer between environments.

RESULTS

Our results suggest that motor skill acquisition of this task occurs at the same rate in both HMD-VR and conventional screen environments. However, the motor skills acquired in HMD-VR did not transfer to the screen environment. While this decrease in motor skill performance when moving to the screen environment was not significantly predicted by self-reported factors, there were trends for correlations with presence and previous HMD-VR experience. Conversely, motor skills acquired in a conventional screen environment not only transferred but improved in HMD-VR, and this increase in motor skill performance could be predicted by self-reported factors of presence, gender, age and video game use.

CONCLUSIONS

These findings suggest that personal factors may predict who is likely to have better transfer of motor skill to and from HMD-VR. Future work should examine whether these and other predictors (i.e., additional personal factors such as immersive tendencies and task-specific factors such as fidelity or feedback) also apply to motor skill transfer from HMD-VR to more dynamic physical environments.

Machine learning and clinical epigenetics: a review of challenges for diagnosis and classification

BACKGROUND

Machine learning is a sub-field of artificial intelligence, which utilises large data sets to make predictions for future events. Although most algorithms used in machine learning were developed as far back as the 1950s, the advent of big data in combination with dramatically increased computing power has spurred renewed interest in this technology over the last two decades.

MAIN BODY

Within the medical field, machine learning is promising in the development of assistive clinical tools for detection of e.g. cancers and prediction of disease. Recent advances in deep learning technologies, a sub-discipline of machine learning that requires less user input but more data and processing power, has provided even greater promise in assisting physicians to achieve accurate diagnoses. Within the fields of genetics and its sub-field epigenetics, both prime examples of complex data, machine learning methods are on the rise, as the field of personalised medicine is aiming for treatment of the individual based on their genetic and epigenetic profiles.

CONCLUSION

We now have an ever-growing number of reported epigenetic alterations in disease, and this offers a chance to increase sensitivity and specificity of future diagnostics and therapies. Currently, there are limited studies using machine learning applied to epigenetics. They pertain to a wide variety of disease states and have used mostly supervised machine learning methods.

The Many Faces of Gene Regulation in Cancer: A Computational Oncogenomics Outlook

Cancer is a complex disease at many different levels. The molecular phenomenology of cancer is also quite rich. The mutational and genomic origins of cancer and their downstream effects on processes such as the reprogramming of the gene regulatory control and the molecular pathways depending on such control have been recognized as central to the characterization of the disease. More important though is the understanding of their causes, prognosis, and therapeutics. There is a multitude of factors associated with anomalous control of gene expression in cancer. Many of these factors are now amenable to be studied comprehensively by means of experiments based on diverse omic technologies. However, characterizing each dimension of the phenomenon individually has proven to fall short in presenting a clear picture of expression regulation as a whole. In this review article, we discuss some of the more relevant factors affecting gene expression control both, under normal conditions and in tumor settings. We describe the different omic approaches that we can use as well as the computational genomic analysis needed to track down these factors. Then we present theoretical and computational frameworks developed to integrate the amount of diverse information provided by such single-omic analyses. We contextualize this within a systems biology-based multi-omic regulation setting, aimed at better understanding the complex interplay of gene expression deregulation in cancer.

An Evaluation of Machine Learning Approaches for the Prediction of Essential Genes in Eukaryotes Using Protein Sequence-Derived Features

The availability of whole-genome sequences and associated multi-omics data sets, combined with advances in gene knockout and knockdown methods, has enabled large-scale annotation and exploration of gene and protein functions in eukaryotes. Knowing which genes are essential for the survival of eukaryotic organisms is paramount for an understanding of the basic mechanisms of life, and could assist in identifying intervention targets in eukaryotic pathogens and cancer. Here, we studied essential gene orthologs among selected species of eukaryotes, and then employed a systematic machine-learning approach, using protein sequence-derived features and selection procedures, to investigate essential gene predictions within and among species. We showed that the numbers of essential gene orthologs comprise small fractions when compared with the total number of orthologs among the eukaryotic species studied. In addition, we demonstrated that machine-learning models trained with subsets of essentiality-related data performed better than random guessing of gene essentiality for a particular species. Consistent with our gene ortholog analysis, the predictions of essential genes among multiple (including distantly-related) species is possible, yet challenging, suggesting that most essential genes are unique to a species. The present work provides a foundation for the expansion of genome-wide essentiality investigations in eukaryotes using machine learning approaches.

SNP variable selection by generalized graph domination

BACKGROUND

High-throughput sequencing technology has revolutionized both medical and biological research by generating exceedingly large numbers of genetic variants. The resulting datasets share a number of common characteristics that might lead to poor generalization capacity. Concerns include noise accumulated due to the large number of predictors, sparse information regarding the p≫n problem, and overfitting and model mis-identification resulting from spurious collinearity. Additionally, complex correlation patterns are present among variables. As a consequence, reliable variable selection techniques play a pivotal role in predictive analysis, generalization capability, and robustness in clustering, as well as interpretability of the derived models.

METHODS AND FINDINGS

K-dominating set, a parameterized graph-theoretic generalization model, was used to model SNP (single nucleotide polymorphism) data as a similarity network and searched for representative SNP variables. In particular, each SNP was represented as a vertex in the graph, (dis)similarity measures such as correlation coefficients or pairwise linkage disequilibrium were estimated to describe the relationship between each pair of SNPs; a pair of vertices are adjacent, i.e. joined by an edge, if the pairwise similarity measure exceeds a user-specified threshold. A minimum k-dominating set in the SNP graph was then made as the smallest subset such that every SNP that is excluded from the subset has at least k neighbors in the selected ones. The strength of k-dominating set selection in identifying independent variables, and in culling representative variables that are highly correlated with others, was demonstrated by a simulated dataset. The advantages of k-dominating set variable selection were also illustrated in two applications: pedigree reconstruction using SNP profiles of 1,372 Douglas-fir trees, and species delineation for 226 grasshopper mouse samples. A C++ source code that implements SNP-SELECT and uses Gurobi optimization solver for the k-dominating set variable selection is available (https://github.com/transgenomicsosu/SNP-SELECT).