Fusing Data Mining,Machine Learning and Traditional Statistics to Detect Biomarkers Associated with Depression

Background

Atheoretical large-scale data mining techniques using machine learning algorithms have promise in the analysis of large epidemiological datasets. This study illustrates the use of a hybrid methodology for variable selection that took account of missing data and complex survey design to identify key biomarkers associated with depression from a large epidemiological study.

Methods

The study used a three-step methodology amalgamating multiple imputation, a machine learning boosted regression algorithm and logistic regression, to identify key biomarkers associated with depression in the National Health and Nutrition Examination Study (2009–2010). Depression was measured using the Patient Health Questionnaire-9 and 67 biomarkers were analysed. Covariates in this study included gender, age, race, smoking, food security, Poverty Income Ratio, Body Mass Index, physical activity, alcohol use, medical conditions and medications. The final imputed weighted multiple logistic regression model included possible confounders and moderators.

Results

After the creation of 20 imputation data sets from multiple chained regression sequences, machine learning boosted regression initially identified 21 biomarkers associated with depression. Using traditional logistic regression methods, including controlling for possible confounders and moderators, a final set of three biomarkers were selected. The final three biomarkers from the novel hybrid variable selection methodology were red cell distribution width (OR 1.15; 95% CI 1.01, 1.30), serum glucose (OR 1.01; 95% CI 1.00, 1.01) and total bilirubin (OR 0.12; 95% CI 0.05, 0.28). Significant interactions were found between total bilirubin with Mexican American/Hispanic group (p = 0.016), and current smokers (p<0.001).

Conclusion

The systematic use of a hybrid methodology for variable selection, fusing data mining techniques using a machine learning algorithm with traditional statistical modelling, accounted for missing data and complex survey sampling methodology and was demonstrated to be a useful tool for detecting three biomarkers associated with depression for future hypothesis generation: red cell distribution width, serum glucose and total bilirubin.     

P2X<Subscript>7</Subscript>R modulation of visually evoked synaptic responses in the retina

P2X₇Rs are distributed throughout all layers of the retina, and thus, their localisation on various cell types puts into question their specific site(s) of action. Using a dark-adapted, ex vivo mouse retinal whole mount preparation, the present study aimed to characterise the effect of P2X₇R activation on light-evoked, excitatory RGC ON-field excitatory post-synaptic potentials (fEPSPs) and on outer retinal electroretinogram (ERG) responses under comparable conditions. The pharmacologically isolated NMDA receptor-mediated RGC ON-fEPSP was reduced in the presence of BzATP, an effect which was significantly attenuated by A438079 and other selective P2X₇R antagonists A804598 or AF27139. In physiological Krebs medium, BzATP induced a significant potentiation of the ERG a-wave, with a concomitant reduction in the b-wave and the power of the oscillatory potentials. Conversely, in the pharmacologically modified Mg²⁺-free perfusate, BzATP reduced both the a-wave and b-wave. The effects of BzATP on the ERG components were suppressed by A438079. A role for P2X₇R function in visual processing in both the inner and outer retina under physiological conditions remains controversial. The ON-fEPSP was significantly reduced in the presence of A804598 but not by A438079 or AF27139. Furthermore, A438079 did not have any effect on the ERG components in physiological Krebs but potentiated and reduced the a-wave and b-wave, respectively, when applied to the pharmacologically modified medium. Therefore, activation of P2X₇Rs affects the function in the retinal ON pathway. The presence of a high concentration of extracellular ATP would most likely contribute to the modulation of visual transmission in the retina in the pathophysiological microenvironment.     

One step purification of alpha(1)-acid glycoprotein from human plasma. Fractionation of its polymorphic allele products

Alpha(1)-acid glycoprotein is a plasma protein that exhibits both microheterogeneity and polymorphism. Its purification from human plasma is usually performed using a sequence of different fractionation steps. Here we report a one-step isolation technique of this protein based upon pseudo-ligand affinity chromatography on immobilized Cibacron Blue F3GA at acidic pH. In addition, the use of two narrow pH elution buffers allows us to separate the two genetic products of this protein, which differ from each other by 21 amino acid substitutions. This technique will facilitate the study of the structural, biological and pharmacokinetic properties of each individual allele product.     

Small shifts in diurnal rhythms are associated with an increase in suicide: The effect of daylight saving

Sleep and Biological Rhythms - Large disruptions of chronobiological rhythms are documented as destabilizing individuals with bipolar disorder; however, the impact of small phase altering events is...     

Determination of nefazodone and its pharmacologically active metabolites in human blood plasma and breast milk by high-performance liquid chromatography

A method is described for the determination of nefazodone and its active metabolites hydroxynefazodone, the dione BMS-180492 and m-chlorophenylpiperazine in blood plasma and expressed human milk based on reversed-phase high-performance liquid chromatography. Measurements were performed on drug-free plasma and expressed human milk spiked with nefazodone and metabolites to prepare and validate standard curves and specimens collected from nursing mothers. Parent drug and metabolites were separated from the biological matrices by solid-phase extraction using CERTIFY columns. Chromatographic separation was achieved with a C₁₈ column and compounds were detected by their absorbance at 205 nm. Trazodone was used as an internal standard. The assay was validated for each analyte in the concentration range 200 to 1200 ng/ml.     

Comorbidity between major depressive disorder and physical diseases: a comprehensive review of epidemiology,mechanisms and management

Populations with common physical diseases – such as cardiovascular diseases, cancer and neurodegenerative disorders – experience substantially higher rates of major depressive disorder (MDD) than the general population. On the other hand, people living with MDD have a greater risk for many physical diseases. This high level of comorbidity is associated with worse outcomes, reduced adherence to treatment, increased mortality, and greater health care utilization and costs. Comorbidity can also result in a range of clinical challenges, such as a more complicated therapeutic alliance, issues pertaining to adaptive health behaviors, drug-drug interactions and adverse events induced by medications used for physical and mental disorders. Potential explanations for the high prevalence of the above comorbidity involve shared genetic and biological pathways. These latter include inflammation, the gut microbiome, mitochondrial function and energy metabolism, hypothalamic-pituitary-adrenal axis dysregulation, and brain structure and function. Furthermore, MDD and physical diseases have in common several antecedents related to social factors (e.g., socioeconomic status), lifestyle variables (e.g., physical activity, diet, sleep), and stressful live events (e.g., childhood trauma). Pharmacotherapies and psychotherapies are effective treatments for comorbid MDD, and the introduction of lifestyle interventions as well as collaborative care models and digital technologies provide promising strategies for improving management. This paper aims to provide a detailed overview of the epidemiology of the comorbidity of MDD and specific physical diseases, including prevalence and bidirectional risk; of shared biological pathways potentially implicated in the pathogenesis of MDD and common physical diseases; of socio-environmental factors that serve as both shared risk and protective factors; and of management of MDD and physical diseases, including prevention and treatment. We conclude with future directions and emerging research related to optimal care of people with comorbid MDD and physical diseases.     

Synthesis of Nucleic Acid Bases by Metal Ferrite Nanoparticles from a Single Carbon Atom Precursor Molecule: Formamide

Origins of Life and Evolution of Biospheres - The synthesis of prebiotic molecules from simple precursors is believed to be a crucial scheme in order to study the origin of life processes. The...     

Castelli risk indexes 1 and 2 are higher in major depression but other characteristics of the metabolic syndrome are not specific to mood disorders

Aims

This study examined whether Castelli risk indexes 1 (total/high-density lipoprotein (HDL) cholesterol) and 2 (low density lipoprotein (LDL)/HDL cholesterol) and other shared metabolic disorders might underpin the pathophysiology of the metabolic syndrome, major depression or bipolar disorder.

Main methods

This cross-sectional study examined 92 major depressed, 49 bipolar depressed and 201 normal controls in whom the Castelli risk indexes 1 and 2 and key characteristics of the metabolic syndrome, i.e. waist/hip circumference, body mass index (BMI), systolic/diastolic blood pressure, total cholesterol, low-density lipoprotein (LDL) and HDL cholesterol, triglycerides, insulin, glucose, hemoglobin A1c (HbA1c) and homocysteine were assessed.

Key findings

Castelli risk indexes 1 and 2 were significantly higher in major depressed patients than in bipolar disorder patients and controls. There were no significant differences in waist or hip circumference, total and LDL cholesterol, triglycerides, plasma glucose, insulin, homocysteine and HbA1c between depression and bipolar patients and controls. Bipolar patients had a significantly higher BMI than major depressed patients and normal controls.

Significance

Major depression is accompanied by increased Castelli risk indexes 1 and 2, which may be risk factors for cardiovascular disease. Other key characteristics of the metabolic syndrome, either metabolic biomarkers or central obesity, are not necessarily specific to major depression or bipolar disorder.     

Chiral determination of mirtazapine in human blood plasma by high-performance liquid chromatography

A method is described for the determination of the two enantiomers of mirtazapine in human blood plasma by high-performance liquid chromatography. Measurements were performed on drug free plasma spiked with mirtazapine and used to prepare and validate standard curves. Levels of enantiomers of mirtazapine were also measured in patients being treated for depression with racemic mirtazapine. Mirtazapine was separated from plasma by solid-phase extraction using CERTIFY columns. Chromatographic separation was achieved using a Chiralpak AD column and pre-column and compounds were detected by their absorption at 290 nm. Imipramine was used as an internal standard. The assay was validated for each analyte in the concentration range 10–100 ng/ml. The coefficient of variance was 16% and 5.5% for(+)-mirtazapine for 10 and 100 ng/ml control specimens respectively and 15% and 7.3% for mirtazapine for 10 and 100 ng/ml control specimens respectively. This assay is appropriate for use in the clinical range. The range of plasma mirtazapine concentrations from eleven patients taking daily doses of 30–45 mg of racemate was <5 to 69 ng/ml for (+)-mirtazapine and 13–88 ng/ml for (−)-mirtazapine for blood specimens collected 10–17.5 h after taking the dose.