SNP interaction detection with random forests in high-dimensional genetic data

ABSTRACT: BACKGROUND: Identifying variants associated with complex human traits in high-dimensional data is a central goal of genome-wide association studies. However, complicated etiologies such as gene-gene interactions are ignored by the univariate analysis usually applied in these studies. Random Forests (RF) are a popular data-mining technique that can accommodate a large number of predictor variables and allow for complex models with interactions. RF analysis produces measures of variable importance that can be used to rank the predictor variables. Thus, single nucleotide polymorphism (SNP) analysis using RFs is gaining popularity as a potential filter approach that considers interactions in high-dimensional data. However, the impact of data dimensionality on the power of RF to identify interactions has not been thoroughly explored. We investigate the ability of rankings from variable importance measures to detect gene-gene interaction effects and their potential effectiveness as filters compared to p-values from univariate logistic regression, particularly as the data becomes increasingly high-dimensional. RESULTS: RF effectively identifies interactions in low dimensional data. As the total number of predictor variables increases, probability of detection declines more rapidly for interacting SNPs than for non-interacting SNPs, indicating that in high-dimensional data the RF variable importance measures are capturing marginal effects rather than capturing the effects of interactions. CONCLUSIONS: While RF remains a promising data-mining technique that extends univariate methods to condition on multiple variables simultaneously, RF variable importance measures fail to detect interaction effects in high-dimensional data in the absence of a strong marginal component, and therefore may not be useful as a filter technique that allows for interaction effects in genome-wide data.     

Enzymological aspects of the pathways for trimethylamine oxidation and C1 assimilation of obligate methylotrophs and restricted facultative methylotrophs.

Extracts of trimethylamine-grown W6A and W3A1 (type M restricted facultative methylotrophs) contain trimethylamine dehydrogenase whereas similar extracts of Bacillus PM6 and Bacillus S2A1 (type L restricted facultative methylotrophs) contain trimethylamine mono-oxygenase and trimethylamine N-oxide demethylase but no trimethylamine dehydrogenase. Extracts of the restricted facultatives and of the obligate methylotroph C2A1 contain hexulose phosphate synthase-hexulose phosphate isomerase activity; hydroxypyruvate reductase was not detected. Neither the restricted facultatives nor the obligates 4B6 and C2A1 contain all the enzymes of the hexulose phosphate cycle of formaldehyde assimilation as originally proposed by Kemp & Quayle (1967). Organisms PM6 and S2A1 lack transaldolase and use a modified cycle involving sedoheptulose 1,7-diphosphate and sedoheptulose diphosphatase. The obligates 4B6 and C2A1, and the type M organisms W6A and W3A1, use a different modification of the assimilatory hexulose phosphate cycle involving the Entner-Doudoroff-pathway enzymes phosphogluconate dehydratase and phospho-2-keto-3-deoxygluconate aldolase. The lack of fructose diphosphate aldolase and hexose diphosphatase in these organisms may be a partial explanation of their restricted growth-substrate range. Enzymological evidence suggests that all the obligates and the restricted facultatives use a dissimilatory hexulose phosphate cycle to accomplish the complete oxidation of formaldehyde to CO2 and water.     

Defective early innate immune response to ectromelia virus in the draining lymph nodes of aged mice due to impaired dendritic cell accumulation

It is known that aging decreases natural resistance to viral diseases due to dysfunctional innate and adaptive immune responses, but the nature of these dysfunctions, particularly in regard to innate immunity, is not well understood. We have previously shown that C57BL/6J (B6) mice lose their natural resistance to footpad infection with ectromelia virus (ECTV) due to impaired maturation and recruitment of natural killer (NK) cells to the draining popliteal lymph node (dLN). More recently, we have also shown that in young B6 mice infected with ECTV, the recruitment of NK cells is dependent on a complex cascade whereby migratory dendritic cells (mDCs) traffic from the skin to the dLN, where they produce CCL2 and CCL7 to recruit inflammatory monocytes (iMOs). In the dLN, mDCs also upregulate NKG2D ligands to induce interferon gamma (IFN‐γ) expression by group 1 innate lymphoid cells (G1‐ILCs), mostly NK in cells but also some ILC1. In response to the IFN‐γ, the incoming uninfected iMOs secret CXCL9 to recruit the critical NK cells. Here, we show that in aged B6 mice, the trafficking of mDCs to the dLN in response to ECTV is decreased, resulting in impaired IFN‐γ expression by G1‐ILCs, reduced accumulation of iMOs, and attenuated CXCL9 production by iMOs, which likely contributes to decrease in NK cell recruitment. Together, these data indicate that defects in the mDC response to viral infection during aging result in a reduced innate immune response in the dLN and contribute to increased susceptibility to viral disease in the aged.     

Effects of lipid peroxidation on adrenal microsomal monooxygenases

Incubation of guinea pig adrenal microsomes with 10^?6 M ferrous (Fe²⁺) ion and adrenal cytosol initiated high levels of lipid peroxidation as measured by the production of malonaldehyde. Cytosol or Fe²⁺ alone had little effect on microsomal malonaldehyde formation. When microsomes were incubated in the presence of Fe²⁺ and cytosol, malonaldehyde levels continued to increase for at least 60 min. Accompanying the lipid peroxidation was a decline in adrenal microsomal monooxygenase activities. The rates of metabolism of xenobiotics (benzphetamine demethylase, benzo[α]pyrene hydroxylase) as well as steroids (21-hydroxylation) decreased as malonaldehyde levels increased. In addition, cytochrome P-450 levels, NADPH- and NADH-cytochrome c reductase activities, and substrate interactions with cytochrome(s) P-450 decreased as lipid peroxidation progressed. Inhibition of lipid peroxidation by increasing microsomal protein concentrations during the incubation period prevented the changes in microsomal metabolism. Malonaldehyde had no direct effects on adrenal microsomal enzyme activities. The results indicate that lipid peroxidation may have significant effects on adrenocortical function, diminishing the capacity for both xenobiotic and steroid metabolism.     

Conversion of spironolactone to an active metabolite in target tissues: formation of 7α-thiospironolactone by microsomal preparations from guinea pig liver,adrenals, kidneys,and testes

Previous studies had established that much of the biological activity of the mineralocorticoid antagonist, spironolactone, depended upon its metabolism, but little was known about the nature of spironolactone metabolites in tissues. A method employing high pressure liquid chromatography has been developed for the separation and mfeasurement of several potential metabolites of spironolactone. Incubation of spironolactone with guinea pig hepatic, adrenal, renal, or testicular microsomal preparations resulted in the production of 7α-thiospironolactone (SC-24813), a compound which is a potent mineralocorticoid antagonist and which promotes the destruction of adrenal and testicular cytochromes P-450. Identification of 7α-thiospironolactone as the tissue metabolite was confirmed by mass spectrometry. The relative rates of production of 7α-thiospironolactone by the microsomal preparations were liver > kidney > adrenal > testis. Spectral data suggest that the effects of spironolactone on adrenal cytochromes P-450 may depend, at least in part, upon its conversion to 7α-thiospironolactone. The results establish that 7α-thiospironolactone is a tissue metabolite of spironolactone and may contribute to the therapeutic actions as well as some of the side effects of the parent drug.     

Primary pneumonic plague in the African Green monkey as a model for treatment efficacy evaluation

Background Primary pneumonic plague is rare among humans, but treatment efficacy may be tested in appropriate animal models under the FDA ‘Animal Rule’. Methods Ten African Green monkeys (AGMs) inhaled 44–255 LD₅₀ doses of aerosolized Yersinia pestis strain CO92. Continuous telemetry, arterial blood gases, chest radiography, blood culture, and clinical pathology monitored disease progression. Results Onset of fever, >39°C detected by continuous telemetry, 52–80 hours post‐exposure was the first sign of systemic disease and provides a distinct signal for treatment initiation. Secondary endpoints of disease severity include tachypnea measured by telemetry, bacteremia, extent of pneumonia imaged by chest x‐ray, and serum lactate dehydrogenase enzyme levels. Conclusions Inhaled Y. pestis in the AGM results in a rapidly progressive and uniformly fatal disease with fever and multifocal pneumonia, serving as a rigorous test model for antibiotic efficacy studies.     

Growth responses of Sphagnum hollows to a growing season lengthening manipulation in Alaskan Arctic tundra

The global increase in surface air temperature has produced an overall lengthening of the growing season by 3–5 days/decade in the Northern Hemisphere during the last 30 years. The direct impact of a longer growing season has not been well documented for Sphagnum moss communities in the Arctic. We hypothesized that an increase in the growing season length may be detrimental to Sphagnum growth as a result of photoinhibition caused by the plants emerging from snow near the seasonal peak of solar irradiance. We conducted an experiment from 1999 to 2002, lengthening the growing season in arctic Alaska, to determine the effects that this simulation of climate change had on the growth of hollows dominated by Sphagnum angustifolium. The lengthened growing season was associated with a decrease in annual moss height increment of 78 and 69 % for 1999 and 2000, respectively. These growth reductions may be related to freeze/thaw episodes and prolonged periods of cold in those years. The growth of individuals exposed to snow removal was also reduced by high global radiation. Overall, snow removal did not significantly affect the seasonal dynamics of growth, but seasonal patterns of growth strongly differed among years. These differences in seasonal dynamics suggest that Sphagnum growth is driven by opportunistic responses to favorable conditions rather than ontogenetic drivers. In addition, we compared environmental variation and growth between control and snow removal plots. Growth of Sphagnum in both treatments was stimulated by warmer soil temperatures and drier conditions. With earlier snowmelt as a result of warmer temperatures, it is likely that S. angustifolium will be subjected to higher levels of radiation and possibly greater photoinhibition which may lead to lower growth rates and significant implications for moss production in tussock tundra.