Improving missing value estimation in microarray data with gene ontology

MOTIVATION: Gene expression microarray experiments produce datasets with frequent missing expression values. Accurate estimation of missing values is an important prerequisite for efficient data analysis as many statistical and machine learning techniques either require a complete dataset or their results are significantly dependent on the quality of such estimates. A limitation of the existing estimation methods for microarray data is that they use no external information but the estimation is based solely on the expression data. We hypothesized that utilizing a priori information on functional similarities available from public databases facilitates the missing value estimation. RESULTS: We investigated whether semantic similarity originating from gene ontology (GO) annotations could improve the selection of relevant genes for missing value estimation. The relative contribution of each information source was automatically estimated from the data using an adaptive weight selection procedure. Our experimental results in yeast cDNA microarray datasets indicated that by considering GO information in the k-nearest neighbor algorithm we can enhance its performance considerably, especially when the number of experimental conditions is small and the percentage of missing values is high. The increase of performance was less evident with a more sophisticated estimation method. We conclude that even a small proportion of annotated genes can provide improvements in data quality significant for the eventual interpretation of the microarray experiments. AVAILABILITY: Java and Matlab codes are available on request from the authors. SUPPLEMENTARY MATERIAL: Available online at http://users.utu.fi/jotatu/GOImpute.html.     

Improving identification of differentially expressed genes by integrative analysis of Affymetrix and Illumina arrays

Together with the widely used Affymetrix microarrays, the recently introduced Illumina platform has become a cost-effective alternative for genome-wide studies. To efficiently use data from both array platforms, there is a pressing need for methods that allow systematic integration of multiple datasets, especially when the number of samples is small. To address these needs, we introduce a meta-analytic procedure for combining Affymetrix and Illumina data in the context of detecting differentially expressed genes between the platforms. We first investigate the effect of different expression change estimation procedures within the platforms on the agreement of the most differentially expressed genes. Using the best estimation methods, we then show the benefits of the integrative analysis in producing reproducible results across bootstrap samples. In particular, we demonstrate its biological relevance in identifying small but consistent changes during T helper 2 cell differentiation.     

Strong seasonality may attenuate trophic cascades: vertebrate predator exclusion in boreal grassland

We studied the indirect effects of vertebrate predator exclusion on plant communities in boreal grassland in western Finland to find out whether the removal of the top trophic level would result in a trophic cascade. Predators were excluded from 1996 to 2000 by eight predator-proof fences (each 0.5 ha) constructed on old fields. Despite a major increase in vole densities, the expected trophic cascade attenuated rapidly so that the indirect effects of predator exclusion were restricted to a few plant species. The cause for the rapid attenuation of the trophic cascade appeared to be strong seasonality, as peak densities of voles were attained at the end of the growing season of vegetation, and vole populations declined before the next growing season so that the herbivory pressure during the growing season remained low or moderate. Accordingly, most plants escaped the heaviest grazing pressure either in time (plants completed their reproduction and withered before winter) or in space (living parts hidden under frozen ground and ice). However, heavy winter herbivory reduced the biomass of available vegetation and killed woody species (willows) at vole peaks, which implies that predator exclusion may have a strong effect on secondary succession. During summer, voles reduced the coverage of only a few preferred food plants ( Elymus repens , Phleum pratense , Vicia cracca ). Voles also maintained annual and biennial species in the community by creating gaps in the closed vegetation. We conclude that abiotic factors (harsh winter conditions) limited peak numbers of herbivores below a threshold density where herbivores could have caused a community-level decline in the biomass of herbaceous plants during summer.     

A field study with geometrid moths to test the coevolution hypothesis of red autumn colours in deciduous trees

Red autumn colouration of trees is the result of newly synthesized anthocyanin pigments in senescing autumn leaves. As anthocyanin accumulation is costly and the trait is not present in all species, anthocyanins must have an adaptive significance in autumn leaves. According to the coevolution hypothesis of autumn colours, red autumn leaves warn herbivorous insects – especially aphids that migrate to reproduce in trees in the autumn – that the tree will not be a suitable host for their offspring in spring due to a high level of chemical defence or lack of nutrients. The signalling allows trees to avoid herbivores and herbivores to choose better host trees. In this study the coevolution hypothesis was tested with four deciduous tree species that have red autumn leaf colouration – European aspen (Populus tremula L.) (Salicaceae), rowan (Sorbus aucuparia L.) (Rosaceae), mountain birch [Betula pubescens ssp. czerepanovii (NI Orlova) Hämet‐Ahti], and dwarf birch (Betula nana L.) (Betulaceae), and with two generalist herbivores, the autumnal moth [Epirrita autumnata (Borkhausen)] and the winter moth [Operophtera brumata (L.)] (both Lepidoptera: Geometridae). Anthocyanin concentrations of autumn leaves were determined from leaf samples and the growth performance parameters of the moth larvae on the study trees were measured in the spring. Trees with higher anthocyanin concentration in the autumn were predicted to be low‐quality food for the herbivores. Our results clearly showed that anthocyanin concentration was not correlated with the growth performance of the moths in any of the studied tree species. Thus, our study does not support the coevolution hypothesis of autumn colours.     

Distinct mechanisms underlay DNA disintegration during apoptosis induced by genotoxic and nongenotoxic agents in neuroblastoma cells

Exposure of mouse NB-2a neuroblastoma cells to genotoxic (etoposide or cytosine arabinoside) or nongenotoxic challenges (serum deprivation or okadaic acid) resulted in progressive cell death with biochemical and morphological characteristics typical of apoptosis. Apoptotic cell death induced by nongenotoxic agents was associated with the disintegration of nuclear DNA into high molecular weight (HMW) and oligonucleosomal-DNA fragments, while the formation of HMW-DNA fragments, but not oligonucleosomal-DNA ladder accompanied apoptosis induced by genotoxic agents. Combination of genotoxic and nongenotoxic insults, i.e. incubation of etoposide-treated cells in the serum-free medium, resulted in an additive effect on the profile of DNA disintegration, which involved both HMW fragmentation pattern as in etoposide alone treated cells and the oligonucleosomal-DNA ladder observed with serum-deprived cells. On the other hand, incubation of serum-deprived cells in the presence of Zn²⁺-ions led to the abrogation of internucleosomal DNA fragmentation but accumulation of HMW-DNA fragments. Differences in the pattern of DNA fragmentation were reproducible in a cell free apoptotic system after treatment of isolated normal nuclei with cytosolic extracts prepared from the cells treated with genotoxic or nogenotoxic apoptotic inducers. Cell free experiments also revealed that activities responsible for the formation of HMW- and oligonucleosomal-DNA fragments are separable in cytosolic extract prepared from the serum-deprived cells. Finally, DNA fragmentation induced by nongenotoxic apoptotic inducers was effectively prevented by cycloheximide and suramin, while both cycloheximide and suramin had only a slight inhibitory effect on DNA fragmentation induced by genotoxic agents. The results presented suggest that distinct pathways underlay disintegration of nuclear DNA during apoptosis induced by genotoxic and nongenotoxic inducers, and that the formation of HMW- and oligonucleosomal-DNA fragments proceeds via separate mechanisms in NB-2a neuroblastoma cells.     

RNA Replication and Membrane Modification Require the Same Functions of Alphavirus Nonstructural Proteins

The alphaviruses induce membrane invaginations known as spherules as their RNA replication sites. Here, we show that inactivation of any function (polymerase, helicase, protease, or membrane association) essential for RNA synthesis also prevents the generation of spherule structures in a Semliki Forest virus trans-replication system. Mutants capable of negative-strand synthesis, including those defective in RNA capping, gave rise to spherules. Recruitment of RNA to membranes in the absence of spherule formation was not detected.