GTI: a novel algorithm for identifying outlier gene expression profiles from integrated microarray datasets

Background

Meta-analysis of gene expression microarray datasets presents significant challenges for statistical analysis. We developed and validated a new bioinformatic method for the identification of genes upregulated in subsets of samples of a given tumour type (‘outlier genes’), a hallmark of potential oncogenes.

Methodology

A new statistical method (the gene tissue index, GTI) was developed by modifying and adapting algorithms originally developed for statistical problems in economics. We compared the potential of the GTI to detect outlier genes in meta-datasets with four previously defined statistical methods, COPA, the OS statistic, the t-test and ORT, using simulated data. We demonstrated that the GTI performed equally well to existing methods in a single study simulation. Next, we evaluated the performance of the GTI in the analysis of combined Affymetrix gene expression data from several published studies covering 392 normal samples of tissue from the central nervous system, 74 astrocytomas, and 353 glioblastomas. According to the results, the GTI was better able than most of the previous methods to identify known oncogenic outlier genes. In addition, the GTI identified 29 novel outlier genes in glioblastomas, including TYMS and CDKN2A. The over-expression of these genes was validated in vivo by immunohistochemical staining data from clinical glioblastoma samples. Immunohistochemical data were available for 65% (19 of 29) of these genes, and 17 of these 19 genes (90%) showed a typical outlier staining pattern. Furthermore, raltitrexed, a specific inhibitor of TYMS used in the therapy of tumour types other than glioblastoma, also effectively blocked cell proliferation in glioblastoma cell lines, thus highlighting this outlier gene candidate as a potential therapeutic target.

Conclusions/Significance

Taken together, these results support the GTI as a novel approach to identify potential oncogene outliers and drug targets. The algorithm is implemented in an R package (Text S1).     

Consistent avoidance of human disturbance over large geographical distances by a migratory bird

Recent work on animal personalities has demonstrated that individuals may show consistent behaviour across situations and contexts. These studies were often carried out in one location and/or during short time intervals. Many animals, however, migrate and spend their life in several geographically distinct locations, and they may either adopt behaviours specific to the local environment or keep consistent behaviours over ecologically distinct locations. Long-distance migratory species offer excellent opportunities to test whether the animals maintain their personalities over large geographical scale, although the practical difficulties associated with these studies have hampered such tests. Here, we demonstrate for the first time consistency in disturbance tolerance behaviour in a long-distance migratory bird, using the common crane Grus grus as an ecological model species. Cranes that hatched in undisturbed habitats in Finland choose undisturbed migratory stop-over sites in Hungary, 1300-2000 km away from their breeding ground. This is remarkable, because these sites are not only separated by large distances, they also differ ecologically: the breeding sites are wooded bogs and subarctic tundra, whereas the migratory stop-over sites are temperate zone alkaline grasslands. The significance of our study goes beyond evolutionary biology and behavioural ecology: local effects on behaviour may carry over large distances, and this hitherto hidden implication of habitat selection needs to be incorporated into conservation planning.     

Body fat measured by a near-infrared interactance device as a predictor of cardiovascular events: the FINRISK'92 cohort

We evaluated how body fat percentage, measured by a portable near-infrared interactance (NIR) device predicts cardiovascular (CVD), coronary heart disease (CHD), and ischemic stroke events in a prospective population-based survey. The study population consisted of 2,842 men and 3,196 women, who participated in the FINRISK'92 survey. Obesity was assessed with BMI, waist circumference, and waist-to-hip ratio (WHR) and body fat percentage measured with an NIR. Mean length of follow-up was 9 years and 3 months. In Cox proportional hazards regression analyses for men, BMI, waist circumference, and WHR as well as body fat percentage were predictors of a CVD event when adjusted for age and for major risk factors. Hazard ratio (HR) per 1 s.d. was 1.27 (95% confidence interval: 1.10-1.48) for body fat percentage, 1.30 (1.16-1.46) for BMI, and 1.31 (1.16-1.50) for waist circumference. Among women, the body fat lost its predictive power in a fully adjusted model. Body fat percentage, BMI, waist circumference, and WHR were predictors of a CHD event both among men and women, whereas body fat percentage did not predict ischemic stroke among either gender. We observed that body fat percentage measured by an NIR device was a significant predictor of CVD and CHD events among men and women, but in our population-based survey, it did not provide any additional predictive power over and above the simpler measures, such as BMI or WHR.     

Polymorphic microsatellite DNA markers in the ant Formica exsecta

Highly polymorphic genetic markers provide a useful tool for estimating important genetic parameters in studies of the evolution of sociality in insects. Here we report 14 polymorphic microsatellite markers developed in the ant Formica exsecta. The number of alleles found ranged between 3 and 18 per locus. These markers were developed for studying genetic population structure and mating structure in F. exsecta populations with varying social organizations (monogyne and polygyne types of societies). Cross‐species amplification indicated that some of the markers might be usable even in species belonging to different subfamilies.     

Nitrogen and Phosphorus in the Finnish Energy System, 1900–2003

In producing power, humans move the nutrients nitrogen (N) and phosphorus (P) from their long‐term geological and biological stocks and release or emit them in soil, water, and the atmosphere. In Finland, peat combustion is an important driver of N and P fluxes from the environment to human economy. The flows of N and P in the Finnish energy system were quantified with partial substance flow analysis, and the driving forces of emissions of nitrogen oxides (NOx) were analyzed using the ImPACT model. In the year 2000 in Finland, 140,000 tonnes of nitrogen entered the energy system, mainly in peat and hard coal. Combustion released an estimated 66,000 tonnes of N as nitrogen oxides (NOx) and nitrous oxides (N2O) and another 74,000 tonnes as elemental N2. Most of the emissions were borne in traffic. At the same time, 6,000 tonnes of P was estimated to enter the Finnish energy system, mostly in peat and wood. Ash was mainly used in earth construction and disposed in landfills; thus negligible levels of P were recycled back to nature. During the twentieth century, fuel‐borne input of N increased 20‐fold, and of P 8‐fold. In 1900–1950, the increasing use of hard coal slowly boosted N input, whereas wood fuels were the main carrier of P. Since 1970, the fluxes have been on the rise. NOx emissions leveled off in the 1980s, though, and then declined in conjunction with improvements in combustion technologies such as NOx removal (de‐NOx) technologies in energy production and catalytic converters in cars.     

Geographical variation in circadian eclosion rhythm and photoperiodic adult diapause inDrosophila littoralis

Summary The time measuring system ofDrosophila littoralis strains originating between 40–70° N was found to be highly variable and latitude dependent. The critical daylength for photoperiodic adult diapause varied from 12 h or no diapause response in the south to 20 h in north. The median timing of pupal eclosion rhythm varied correspondingly from 21 h to 12 h from lights off in LD 321, and the period of free-running rhythm of eclosion from 24 h to 19 h. The phase of the free-running rhythm was also variable, and correlated with the phase of the entrained rhythm. Latitudinal variation in the entrained rhythm of eclosion and in diapause is adaptive, leading to eclosion early in the morning and to overwintering at the adult stage. In some strains with a late phase of eclosion, strong transient cycles were seen following the transition from LL to DD. A total damping of the free-running eclosion rhythm within 2–7 days was common to all strains. This damping was more pronounced in the northern strains. The phase and period of eclosion rhythms were statistically independent. Diapause was not correlated with any parameters of the eclosion rhythm in the analysis. Diapause may still be influenced by the period of the eclosion rhythm, even though its minor contribution may be masked by a more variable, eclosion rhythm independent system in the determination of diapause.Abbreviations, symbols and terms LD Light/dark; as in LD 321 meaning a cycle of 3 h light21 h darkness - LL Continuous light - DD Continuous darkness - T Period of a Zeitgeber cycle - Natural period of eclosion rhythm in constant conditions - _EL Phase of the free-running rhythm of eclosion - A Amplitude of the free-running rhythm of eclosion; possible range is from 4.17% (no rhythmicity) to 20% (the daily eclosion peaks 2–6 within 5 h each) - P Persistence of the free-running rhythm of eclosion; the number of daily eclosion peaks where the mean for five highest hourly percentages still exceed 6% - A phase shift, expressed in h; a re-setting of a rhythm; either as an advance shift (i.e. earlier= +), or as a delay shift (i.e. later = –) - PRC Phase-response curve - _LD Phase of entrained rhythm of eclosion; e.g. _{LD 321} is the median hour of eclosion peak from lights off at LD 321 - SD _ecl Amplitude of the entrained rhythm of eclosion; the smaller SD_ecl the higher the amplitude - PPRC Photoperiodic response curve; proportion of females in diapause displayed as a function of daylength - CDL Critical photoperiod; the photoperiod in the 24 h LD cycle at which 50% of the population studied diapauses - SD _diap Accuracy of diapause response of a strain; the smaller the SD_diap the more accurate the response - Cdl The main locus controlling CDL inD. littoralis     

Strain while skiing and hauling a sledge or carrying a backpack

Eight soldiers on skis transported three loads of different weights on the level, uphill and downhill. The load was placed either on a cargo sledge or in a backpack or divided between the sledge and the backpack. The sledge had a new type of haulage-shaft, which was fixed to both sides of the pelvis. A service belt spread the pull over the whole upper body. The physical stain of different transport methods and the serviceability of the sledge was studied by measuring heart rate (HR), oxygen consumption, ventilation, and perceived exertion. The results indicate that both absolute and relative strain were systematically lower when pulling the load on the sledge than when carrying it in the backpack and on the sledge. HR when pulling a load equal to the human body weight on the sledge was on average 133 beats X min-1; HR was significantly higher 144 beats X min-1 when the load was divided between backpack and sledge. At the lower load level the differences between the transport methods were not significant for HR, oxygen consumption or ventilation. Uphill travel increased oxygen consumption by about 50% over that on the level. Perceived exertion at all load levels was significantly lower with the sledge than with the backpack alone or in combination. The estimated maximal allowable working time emphasized the advantage of the sledge and the importance of high physical working capacity. The maneuverability of the sledge with the new haulage shaft was good and the braking mechanisms worked well.(ABSTRACT TRUNCATED AT 250 WORDS)     

Training effects of cross-country skiing and running on maximal aerobic cycle performance and on blood lipids

Two experiments were carried out to compare the cardiorespiratory and metabolic effects of cross-country skiing and running training during two successive winters. Forty-year-old men were randomly assigned into skiing (n = 15 in study 1, n = 16 in study 2), running (n = 16 in study 1 and n = 16 in study 2) and control (n = 17 in study 1 and n = 16 in study 2) groups. Three subjects dropped out of the programme. The training lasted 9-10 weeks with 40-min exercise sessions three times each week. The training intensity was controlled at 75%-85% of the maximal oxygen consumption (VO2max) using portable heart rate metres and the mean heart rate was 156-157 beats.min-1 in the training groups. In the pooled data of the two studies the mean increase in the VO2max (in ml.min-1.kg-1) on a cycle ergometer was 17% for the skiing group, 13% for the running group and 2% for the control group. The increase in VO2max was highly significant in the combined exercise group compared to the control group but did not differ significantly between the skiing and running groups. The fasting serum concentrations of lipoproteins and insulin did not change significantly in any of the groups. These results suggested that training by cross-country skiing and running of the same duration and intensity at each session for 9-10 weeks improved equally the cardiorespiratory fitness of untrained middle-aged men.