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.     

In Search of Functional Advantages of Knots in Proteins

We analysed the structure of deeply knotted proteins representing three unrelated families of knotted proteins. We looked at the correlation between positions of knotted cores in these proteins and such local structural characteristics as the number of intra-chain contacts, structural stability and solvent accessibility. We observed that the knotted cores and especially their borders showed strong enrichment in the number of contacts. These regions showed also increased thermal stability, whereas their solvent accessibility was decreased. Interestingly, the active sites within these knotted proteins preferentially located in the regions with increased number of contacts that also have increased thermal stability and decreased solvent accessibility. Our results suggest that knotting of polypeptide chains provides a favourable environment for the active sites observed in knotted proteins. Some knotted proteins have homologues without a knot. Interestingly, these unknotted homologues form local entanglements that retain structural characteristics of the knotted cores.     

Two Independent Functions of Collier/Early B Cell Factor in the Control of Drosophila Blood Cell Homeostasis

Blood cell production in the Drosophila hematopoietic organ, the lymph gland, is controlled by intrinsic factors and extrinsic signals. Initial analysis of Collier/Early B Cell Factor function in the lymph gland revealed the role of the Posterior Signaling Center (PSC) in mounting a dedicated cellular immune response to wasp parasitism. Further, premature blood cell differentiation when PSC specification or signaling was impaired, led to assigning the PSC a role equivalent to the vertebrate hematopoietic niche. We report here that Collier is expressed in a core population of lymph gland progenitors and cell autonomously maintains this population. The PSC contributes to lymph gland homeostasis by regulating blood cell differentiation, rather than by maintaining core progenitors. In addition to PSC signaling, switching off Collier expression in progenitors is required for efficient immune response to parasitism. Our data show that two independent sites of Collier/Early B Cell Factor expression, hematopoietic progenitors and the PSC, achieve control of hematopoiesis.     

Oral Administration of Polymyxin B Modulates the Activity of Lipooligosaccharide E. coli B against Lung Metastases in Murine Tumor Models

Introduction

Polymyxin B (PmB) belongs to the group of cyclic peptide antibiotics, which neutralize the activity of LPS by binding to lipid A. The aim of this study was to analyze the effect of PmB on the biological activity of lipooligosaccharide (LOS E. coli B,rough form of LPS) in vitro and in experimental metastasis models.

Results

Cultures of murine macrophage J774A.1 cells and murine bone marrow-derived dendritic cells (BM-DC) stimulated in vitro with LOS and supplemented with PmB demonstrated a decrease in inflammatory cytokine production (IL-6, IL-10, TNF-α) and down-regulation of CD40, CD80, CD86 and MHC class II molecule expression. Additionally, PmB suspended in drinking water was given to the C57BL/6 mice seven or five days prior to the intravenous injection of B16 or LLC cells and intraperitoneal application of LOS. This strategy of PmB administration was continued throughout the duration of the experiments (29 or 21 days). In B16 model, statistically significant decrease in the number of metastases in mice treated with PmB and LOS (p<0.01) was found on the 14th day of the experiments, whereas the most intensive changes in surface-antigen expression and ex vivo production of IL-6, IL-1β and TNF-α by peritoneal cells were observed 7 days earlier. By contrast, antigen expression and ex vivo production of IL-6, IL-10, IFN-γ by splenocytes remained relatively high and stable. Statistically significant decrease in LLC metastases number was observed after the application of LOS (p<0.01) and in the group of mice preconditioned by PmB and subsequently treated with LOS (LOS + PmB, p<0.01).

Conclusions

In conclusion, prolonged in vivo application of PmB was not able to neutralize the LOS-induced immune cell activity but its presence in the organism of treated mice was important in modulation of the LOS-mediated response against the development of metastases.     

CACNA1H missense mutations associated with amyotrophic lateral sclerosis alter Cav3.2 T-type calcium channel activity and reticular thalamic neuron firing

Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease that affects nerve cells in the brain and the spinal cord. In a recent study by Steinberg and colleagues, 2 recessive missense mutations were identified in the Ca_v3.2 T-type calcium channel gene (CACNA1H), in a family with an affected proband (early onset, long duration ALS) and 2 unaffected parents. We have introduced and functionally characterized these mutations using transiently expressed human Ca_v3.2 channels in tsA-201 cells. Both of these mutations produced mild but significant changes on T-type channel activity that are consistent with a loss of channel function. Computer modeling in thalamic reticular neurons suggested that these mutations result in decreased neuronal excitability of thalamic structures. Taken together, these findings implicate CACNA1H as a susceptibility gene in amyotrophic lateral sclerosis.     

Modulation of Cav3.2 T-type calcium channel permeability by asparagine-linked glycosylation

Low-voltage-gated T-type calcium channels are expressed throughout the nervous system where they play an essential role in shaping neuronal excitability. Defects in T-type channel expression have been linked to various neuronal disorders including neuropathic pain and epilepsy. Currently, little is known about the cellular mechanisms controlling the expression and function of T-type channels. Asparagine-linked glycosylation has recently emerged as an essential signaling pathway by which the cellular environment can control expression of T-type channels. However, the role of N-glycans in the conducting function of T-type channels remains elusive. In the present study, we used human Ca_v3.2 glycosylation-deficient channels to assess the role of N-glycosylation on the gating of the channel. Patch-clamp recordings of gating currents revealed that N-glycans attached to hCa_v3.2 channels have a minimal effect on the functioning of the channel voltage-sensor. In contrast, N-glycosylation on specific asparagine residues may have an essential role in the conducting function of the channel by enhancing the channel permeability and / or the pore opening of the channel. Our data suggest that modulation of N-linked glycosylation of hCa_v3.2 channels may play an important physiological role, and could also support the alteration of T-type currents observed in disease states.     

Risk Factors for Normal and High-Tension Glaucoma in Poland in Connection with Polymorphisms of the Endothelial Nitric Oxide Synthase Gene

Aim

The purpose of this study was to evaluate the influence of polymorphisms of the eNOS gene on the clinical status of patients with normal and high tension glaucoma.

Methods

266 Polish Caucasian patients with primary open angle glaucoma were studied. Of the 266, 156 had normal tension glaucoma (NTG) and 110 high tension glaucoma (HTG). DNA material was isolated from peripheral venous blood using commercial kits. Real-time PCR reaction was used to amplify the promoter site of the endothelial nitric oxide synthase (eNOS) gene, including the single nucleotide polymorphism (SNP) site T-786C and part of the 7^th exon of eNOS, including G894T SNP. Genotypes were determined with TaqMan SNP Genotyping Assays.

Results

There were no significant differences in frequencies of the allelic variants of both polymorphisms. In G894T SNP, however, the wild GG form was more common in the HTG group. The SNP of the eNOS gene did not significantly influence the progression rate in either of the groups studied. There were no differences in variants of the eNOS gene regarding the necessity for and success of surgery and the progression of the disease. In the NTG group, no statistical correlation was observed between G894T, T786C polymorphism variants, and risk factors such as optic disc haemorrhages, optic disc notches, and peripapillary atrophy. Mean diastolic and systolic pressure during the day and night were lowest in NTG patients with the CC variant of the T786C polymorphism. No statistical correlation was observed between the G894T and T786C polymorphisms and capillaroscopic examination results.

Conclusions

Genotype frequencies are similar for both the eNOS G894T and T-786C polymorphisms in NTG and HTG patients. These polymorphisms do not correlate with risk factors and do not influence the state of the capillary system in NTG patients. Systolic blood pressure is lower in NTG patients with mutated alleles of both polymorphisms.     

The Relationships between Polymorphisms in Genes Encoding the Growth Factors TGF-β1, PDGFB,EGF, bFGF and VEGF-A and the Restenosis Process in Patients with Stable Coronary Artery Disease Treated with Bare Metal Stent

Background

Neointima forming after stent implantation consists of vascular smooth muscle cells (VSMCs) in 90%. Growth factors TGF-β1, PDGFB, EGF, bFGF and VEGF-A play an important role in VSMC proliferation and migration to the tunica intima after arterial wall injury. The aim of this paper was an analysis of functional polymorphisms in genes encoding TGF-β1, PDGFB, EGF, bFGF and VEGF-A in relation to in-stent restenosis (ISR).

Materials and Methods

265 patients with a stable coronary artery disease (SCAD) hospitalized in our center in the years 2007–2011 were included in the study. All patients underwent stent implantation at admission to the hospital and had another coronary angiography performed due to recurrence of the ailments or a positive result of the test assessing the coronary flow reserve. Angiographically significant ISR was defined as stenosis >50% in the stented coronary artery segment. The patients were divided into two groups–with angiographically significant ISR (n = 53) and without significant ISR (n = 212). Additionally, the assessment of late lumen loss (LLL) in vessel was performed. EGF rs4444903 polymorphism was genotyped using the PCR-RFLP method whilst rs1800470 (TGFB1), rs2285094 (PDGFB) rs308395 (bFGF) and rs699947 (VEGF-A) were determined using the TaqMan method.

Results

Angiographically significant ISR was significantly less frequently observed in the group of patients with the A/A genotype of rs1800470 polymorphism (TGFB1) versus patients with A/G and G/G genotypes. In the multivariable analysis, LLL was significantly lower in patients with the A/A genotype of rs1800470 (TGFB1) versus those with the A/G and G/G genotypes and higher in patients with the A/A genotype of the VEGF-A polymorphism versus the A/C and C/C genotypes. The C/C genotype of rs2285094 (PDGFB) was associated with greater LLL compared to C/T heterozygotes and T/T homozygotes.

Conclusions

The polymorphisms rs1800470, rs2285094 and rs6999447 of the TGFB1, PDGFB and VEGF-A genes, respectively, are associated with LLL in patients with SCAD treated by PCI with a metal stent implantation.     

The Impact of Short-Term,Intensive Antifolate Treatment (with Pyrimethamine and Sulfadoxine) and Antibiotics Followed by Long-Term,Secondary Antifolate Prophylaxis on the Rate of Toxoplasmic Retinochoroiditis Recurrence

PurposeTo assess the impact of intensive antifolate treatment, followed by secondary antifolate prophylaxis (A-SP) on the recurrence rate of toxoplasmic retinochoroiditis (TRC). To investigate whether there are any other factors potentially predisposing for recurrence.ResultsWhen secondary antifolate prophylaxis (A-SP) was instituted immediately after the treatment for TRC, the probability of 3-year recurrence–free survival after the first course of A-SP was 90.9%. A recurrence was most likely approximately 3.5 years after the first treatment. A univariate Cox regression model demonstrated that a risk for recurrence was 2.82 times higher (p = 0.02) in patients with retinal scars. In the multivariate analysis, the risk for recurrence was 2.41 higher (p = 0.06). In patients with haemorrhagic lesions the risk for recurrence was lower, aRR = 0.17 (approaching borderline statistical significance p = 0.08).ConclusionsWith the institution of A-SP of immediately after the intensive treatment for TRC, i.e. when a reactivation was most likely, there was no recurrence during A-SP. Following A-SP the recurrence rates were low and recurrence-free periods tended to be longer. The treatment regimen employed had a beneficial effect on the recurrence interval as it reduced and delayed the highest probability of recurrence.