Batch effect confounding leads to strong bias in performance estimates obtained by cross-validation.

BACKGROUND: With the large amount of biological data that is currently publicly available, many investigators combine multiple data sets to increase the sample size and potentially also the power of their analyses. However, technical differences ("batch effects") as well as differences in sample composition between the data sets may significantly affect the ability to draw generalizable conclusions from such studies. FOCUS: The current study focuses on the construction of classifiers, and the use of cross-validation to estimate their performance. In particular, we investigate the impact of batch effects and differences in sample composition between batches on the accuracy of the classification performance estimate obtained via cross-validation. The focus on estimation bias is a main difference compared to previous studies, which have mostly focused on the predictive performance and how it relates to the presence of batch effects. DATA: We work on simulated data sets. To have realistic intensity distributions, we use real gene expression data as the basis for our simulation. Random samples from this expression matrix are selected and assigned to group 1 (e.g., 'control') or group 2 (e.g., 'treated'). We introduce batch effects and select some features to be differentially expressed between the two groups. We consider several scenarios for our study, most importantly different levels of confounding between groups and batch effects. METHODS: We focus on well-known classifiers: logistic regression, Support Vector Machines (SVM), k-nearest neighbors (kNN) and Random Forests (RF). Feature selection is performed with the Wilcoxon test or the lasso. Parameter tuning and feature selection, as well as the estimation of the prediction performance of each classifier, is performed within a nested cross-validation scheme. The estimated classification performance is then compared to what is obtained when applying the classifier to independent data.

Safety and accuracy of robot-assisted versus fluoroscopy-guided pedicle screw insertion for degenerative diseases of the lumbar spine: a matched cohort comparison.

Object Recent years have been marked by efforts to improve the quality and safety of pedicle screw placement in spinal instrumentation. The aim of the present study is to compare the accuracy of the SpineAssist robot system with conventional fluoroscopy-guided pedicle screw placement. Methods Ninety-five patients suffering from degenerative disease and requiring elective lumbar instrumentation were included in the study. The robot cohort (Group I; 55 patients, 244 screws) consisted of an initial open robot-assisted subgroup (Subgroup IA; 17 patients, 83 screws) and a percutaneous cohort (Subgroup IB, 38 patients, 161 screws). In these groups, pedicle screws were placed under robotic guidance and lateral fluoroscopic control. In the fluoroscopy-guided cohort (Group II; 40 patients, 163 screws) screws were inserted using anatomical landmarks and lateral fluoroscopic guidance. The primary outcome measure was accuracy of screw placement on the Gertzbein-Robbins scale (Grade A to E and R [revised]). Secondary parameters were duration of surgery, blood loss, cumulative morphine, and length of stay. Results In the robot group (Group I), a perfect trajectory (A) was observed in 204 screws (83.6%). The remaining screws were graded B (n = 19 [7.8%]), C (n = 9 [3.7%]), D (n = 4 [1.6%]), E (n = 2 [0.8%]), and R (n = 6 [2.5%]). In the fluoroscopy-guided group (Group II), a completely intrapedicular course graded A was found in 79.8% (n = 130). The remaining screws were graded B (n = 12 [7.4%]), C (n = 10 [6.1%]), D (n = 6 [3.7%]), and E (n = 5 [3.1%]). The comparison of "clinically acceptable" (that is, A and B screws) was neither different between groups (I vs II [p = 0.19]) nor subgroups (Subgroup IA vs IB [p = 0.81]; Subgroup IA vs Group II [p = 0.53]; Subgroup IB vs Group II [p = 0.20]). Blood loss was lower in the robot-assisted group than in the fluoroscopy-guided group, while duration of surgery, length of stay, and cumulative morphine dose were not statistically different. Conclusions Robot-guided pedicle screw placement is a safe and useful tool for assisting spine surgeons in degenerative spine cases. Nonetheless, technical difficulties remain and fluoroscopy backup is advocated.

Effect of conditioned media, nutritive elements, and mitotic synchronization on the accuracy of the cytogenetic analysis in acute nonlymphocytic leukemia patients presenting with inv(16)/t(16;16) or t(15;17).

To improve the yield of the cytogenetic analysis in patients with acute nonlymphocytic leukemia (ANLL), six culture conditions for bone marrow or peripheral blood cells were tested in parallel. Two conditioned media (CM), phytohemagglutinin leukocyte PHA-LCM and 5637 CM, nutritive elements (NE), and methotrexate (MTX) cell synchronization were investigated in 14 patients presenting with either inv(16)/ t(16;16) (group 1, n = 9 patients) or t(15;17) (group 2, n = 5). The criteria used to identify the most favorable culture conditions were the mitotic index (MI), the morphological index (MorI), and the percentage of abnormal metaphases. In the presence of PHA-LCM and 5637 CM, the MI were significantly increased in group 2, whereas in the MTX conditions, MI remained very low in both groups. The values of the MorI did not reveal any significant changes in chromosome resolution between the conditions in either group. The addition of NE did not have a positive effect in quantity or quality of metaphases. Because of the variability of the response of leukemic cells to different stimulations in vitro, several culture conditions in parallel are required to ensure a satisfactory yield of the chromosome analysis in ANLL.

Melanin-based coloration covaries with ovary size in an age-specific manner in the barn owl.

While the adaptive function of black eumelanin-based coloration is relatively well known, the function of reddish-brown pheomelanin-based coloration is still unclear. Only a few studies have shown or suggested that the degree of reddish-brownness is associated with predator-prey relationships, reproductive parameters, growth rate and immunity. To gain insight into the physiological correlates of melanin-based coloration, I collected barn owl (Tyto alba) cadavers and examined the covariation between this colour trait and ovary size, an organ that increases in size before reproduction. A relationship is expected because melanin-based coloration often co-varies with sexual activity. The results showed that reddish-brown juveniles had larger ovaries than whiter juveniles particularly in individuals in poor condition and outside the breeding season, while in birds older than 2 years lightly coloured females had larger ovaries than reddish-brown conspecifics. As barn owls become less reddish-brown between the first and second year of age, the present study suggests that reddish-brown pheomelanic and whitish colorations are associated with juvenile- and adult-specific adaptations, respectively.

Population density and optimal body size.

The population density of an organism is one of the main aspects of its environment, and shoud therefore strongly influence its adaptive strategy. The r/K theory, based on the logistic model, was developed to formalize this influence. K-selectioon is classically thought to favour large body sizes. This prediction, however, cannot be directly derived from the logistic model: some auxiliary hypotheses are therefor implicit. These are to be made explicit if the theory is to be tested. An alternative approach, based on the Euler-Lotka equation, shows that density itself is irrelevant, but that the relative effect of density on adult and juvenile features is crucial. For instance, increasing population will select for a smaller body size if the density affects mainly juvenile growth and/or survival. In this case, density shoud indeed favour large body sizes. The theory appears nevertheless inconsistent, since a probable consequence of increasing body size will be a decrease in the carrying capacity

Extra-pair paternity, testes size and testosterone level in relation to colour polymorphism in the barn owl Tyto alba

In many bird populations, individuals display one of several genetically inherited colour morphs. Colour polymorphism can be maintained by several mechanisms one of which being frequency-dependent selection with colour morphs signalling alternative mating strategies. One morph may be dominant and territorial, and another one adopt a sneaky behaviour to gain access to fertile females. We tested this hypothesis in the barn owl Tyto alba in which coloration varies from reddish-brown to white. This trait is heritable and neither sensitive to the environment in which individuals live nor to body condition. In Switzerland, reddish-brown males were observed to feed their brood at a higher rate and to produce more offspring than white males. This observation lead us to hypothesize that white males may equalise fitness by investing more effort in extra-pair copulations. This hypothesis predicts that lighter Coloured males produce more extra-pair young, have larger testes and higher levels of circulating testosterone. However, our results are not consistent with these three predictions. First, paternity analyses of 54 broods with a total of 211 offspring revealed that only one young was not sired by the male that was feeding it. Second, testes size was not correlated with male plumage coloration suggesting that white males are not sexually more active. Finally, in nestlings at the time of feather growth testosterone level was not related to plumage coloration suggesting that this androgen is not required for the expression of this plumage trait. Our study therefore indicates that in the barn owl colour polymorphism plays no role in the probability of producing extra-pair young.

The ontogeny of sexual size dimorphism of a moth: when do males and females grow apart?

Sexual dimorphism in body size (sexual size dimorphism) is common in many species. The sources of selection that generate the independent evolution of adult male and female size have been investigated extensively by evolutionary biologists, but how and when females and males grow apart during ontogeny is poorly understood. Here we use the hawkmoth, Manduca sexta, to examine when sexual size dimorphism arises by measuring body mass every day during development. We further investigated whether environmental variables influence the ontogeny of sexual size dimorphism by raising moths on three different diet qualities (poor, medium and high). We found that size dimorphism arose during early larval development on the highest quality food treatment but it arose late in larval development when raised on the medium quality food. This female-biased dimorphism (females larger) increased substantially from the pupal-to-adult stage in both treatments, a pattern that appears to be common in Lepidopterans. Although dimorphism appeared in a few stages when individuals were raised on the poorest quality diet, it did not persist such that male and female adults were the same size. This demonstrates that the environmental conditions that insects are raised in can affect the growth trajectories of males and females differently and thus when dimorphism arises or disappears during development. We conclude that the development of sexual size dimorphism in M. sexta occurs during larval development and continues to accumulate during the pupal/adult stages, and that environmental variables such as diet quality can influence patterns of dimorphism in adults.

The red island and the seven dwarfs: body size reduction in Cheirogaleidae

AimSmall body size in Madagascar's dwarf and mouse lemurs (Cheirogaleidae) is generally viewed as primitive. We investigated the evolution of body size in this family and in its sister-taxon, the Lepilemuridae, from phylogenetic, ontogenetic and adaptive perspectives. LocationMadagascar. MethodsWe used a phylogenetic method to reconstruct the evolution of body size in lemurs, and allometric regression models of gestation periods and static and growth allometries in Cheirogaleidae and Lepilemuridae to test the hypothesis that dwarfing occurred as a result of truncated ontogeny (progenesis). We also examined adaptive hypotheses relating body size to environmental variability, life history, seasonality of reproduction, hypothermy (use of torpor), and a diet rich in plant exudates. ResultsOur results indicated that cheirogaleids experienced at least four independent events of body size reduction from an ancestor as large as living Lepilemuridae, by means of progenesis. Our interpretation is supported by the paedomorphic appearance and parallel ontogenetic trajectories of the dwarf taxa, as well as their very short gestation periods and increased fecundity. Lepilemur species that occupy more predictable environments are significantly larger than those occupying unpredictable habitats. Main conclusionsCheirogaleidae appear to be paedomorphic dwarfs, a consequence of progenesis, probably as an adaptation to high environmental unpredictability. Although the capacity to use hypothermy is related to small body size, this advantage is unlikely to have driven dwarfing in cheirogaleids. We propose that gummmivory/exudativory co-evolved with body size reduction in this clade, probably from a folivorous ancestor. Their small size is derived, and their suitability as models for the ancestral primate' is therefore dubious.

Clutch size and malarial parasites in female great tits.

Life-history models predict an evolutionary trade-off in the allocation of resources to current versus future reproduction. This corresponds, at the physiological level, to a trade-off in the allocation of resources to current reproduction or to the immune system, which will enhance survival and therefore future reproduction. For clutch size, life-history models predict a positive correlation between current measurement in eggs and the subsequent parasite load. Tn a population of great tits, we analyzed the correlation between natural clutch size of females and the subsequent prevalence of Plasmodium spp., a potentially harmful blood parasite. Females that showed, 14 days after hatching of the nestlings, an infection with Plasmodium had a significantly larger clutch (9.3 eggs +/- 0.5 SE, n = 18) than uninfected females (8.0 eggs +/- 0.2 SE, n = 80), as predicted by the allocation trade-off. Clutch size was positively correlated with tile prevalence of Plasmodium, but brood size 14 days after hatching was not. This suggests that females incur higher costs during laying the clutch than during rearing nestlings. Infection status of some females changed between years, and these changes were significantly correlated with a change in clutch size as predicted by die trade-off. The link between reproductive effort and parasitism may represent a possible mechanism by which the cost of egg production is mediated into future survival and may thereby be an important selective force in the shaping of clutch size.

A program for the computation of power and determination of sample size in hierarchical experimental designs.

We have devised a program that allows computation of the power of F-test, and hence determination of appropriate sample and subsample sizes, in the context of the one-way hierarchical analysis of variance with fixed effects. The power at a fixed alternative is an increasing function of the sample size and of the subsample size. The program makes it easy to obtain the power of F-test for a range of values of sample and subsample sizes, and therefore the appropriate sizes based on a desired power. The program can be used for the 'ordinary' case of the one-way analysis of variance, as well as for hierarchical analysis of variance with two stages of sampling. Examples are given of the practical use of the program.

Dpp signaling activity requires Pentagone to scale with tissue size in the growing Drosophila wing imaginal disc.

The wing of the fruit fly, Drosophila melanogaster, with its simple, two-dimensional structure, is a model organ well suited for a systems biology approach. The wing arises from an epithelial sac referred to as the wing imaginal disc, which undergoes a phase of massive growth and concomitant patterning during larval stages. The Decapentaplegic (Dpp) morphogen plays a central role in wing formation with its ability to co-coordinately regulate patterning and growth. Here, we asked whether the Dpp signaling activity scales, i.e. expands proportionally, with the growing wing imaginal disc. Using new methods for spatial and temporal quantification of Dpp activity and its scaling properties, we found that the Dpp response scales with the size of the growing tissue. Notably, scaling is not perfect at all positions in the field and the scaling of target gene domains is ensured specifically where they define vein positions. We also found that the target gene domains are not defined at constant concentration thresholds of the downstream Dpp activity gradients P-Mad and Brinker. Most interestingly, Pentagone, an important secreted feedback regulator of the pathway, plays a central role in scaling and acts as an expander of the Dpp gradient during disc growth.

Genetic differentiation for size at first reproduction through male versus female functions in the widespread Mediterranean tree Pinus pinaster.

BACKGROUND AND AIMS: The study of local adaptation in plant reproductive traits has received substantial attention in short-lived species, but studies conducted on forest trees are scarce. This lack of research on long-lived species represents an important gap in our knowledge, because inferences about selection on the reproduction and life history of short-lived species cannot necessarily be extrapolated to trees. This study considers whether the size for first reproduction is locally adapted across a broad geographical range of the Mediterranean conifer species Pinus pinaster. In particular, the study investigates whether this monoecious species varies genetically among populations in terms of whether individuals start to reproduce through their male function, their female function or both sexual functions simultaneously. Whether differences among populations could be attributed to local adaptation across a climatic gradient is then considered. METHODS: Male and female reproduction and growth were measured during early stages of sexual maturity of a P. pinaster common garden comprising 23 populations sampled across the species range. Generalized linear mixed models were used to assess genetic variability of early reproductive life-history traits. Environmental correlations with reproductive life-history traits were tested after controlling for neutral genetic structure provided by 12 nuclear simple sequence repeat markers. KEY RESULTS: Trees tended to reproduce first through their male function, at a size (height) that varied little among source populations. The transition to female reproduction was slower, showed higher levels of variability and was negatively correlated with vegetative growth traits. Several female reproductive traits were correlated with a gradient of growth conditions, even after accounting for neutral genetic structure, with populations from more unfavourable sites tending to commence female reproduction at a lower individual size. CONCLUSIONS: The study represents the first report of genetic variability among populations for differences in the threshold size for first reproduction between male and female sexual functions in a tree species. The relatively uniform size at which individuals begin reproducing through their male function probably represents the fact that pollen dispersal is also relatively invariant among sites. However, the genetic variability in the timing of female reproduction probably reflects environment-dependent costs of cone production. The results also suggest that early sex allocation in this species might evolve under constraints that do not apply to other conifers.

Male body size and breeding tubercles are both linked to intra-sexual dominance and reproductive success in the minnow

Male dominance hierarchies are usually linked to relative body size and to weapon size, that is, to determinants of fighting ability. Secondary sexual characters that are not directly used as weapons could still be linked to dominance if they reveal determination or overall health and vigour and hence, indirectly, fighting ability. We studied the mating behaviour of the minnow, Phoxinus phoxinus, a cyprinid fish in which males develop breeding tubercles during the spawning season. The function of these breeding tubercles is still not clear. Using microsatellite markers, we determined male reproductive success under controlled conditions. The minnows were territorial and quickly established a dominance hierarchy at the beginning of the spawning season. Dominance was strongly and positively linked to fertilization success. Although body size and number of breeding tubercles were not significantly correlated in our sample, both large males and males with many breeding tubercles were more dominant and achieved higher fertilization success than small males or males with few tubercles. We found multimale fertilization in most clutches, suggesting that sperm competition is important in this species. Females showed behaviour that may be linked to spawning decision, that is, male dominance might not be the only determinant of male reproductive success in minnows