miR-7 controls glutamatergic transmission and neuronal connectivity in a Cdr1as-dependent manner.

The circular RNA (circRNA) Cdr1as is conserved across mammals and highly expressed in neurons, where it directly interacts with microRNA miR-7. However, the biological function of this interaction is unknown. Here, using primary cortical murine neurons, we demonstrate that stimulating neurons by sustained depolarization rapidly induces two-fold transcriptional upregulation of Cdr1as and strong post-transcriptional stabilization of miR-7. Cdr1as loss causes doubling of glutamate release from stimulated synapses and increased frequency and duration of local neuronal bursts. Moreover, the periodicity of neuronal networks increases, and synchronicity is impaired. Strikingly, these effects are reverted by sustained expression of miR-7, which also clears Cdr1as molecules from neuronal projections. Consistently, without Cdr1as, transcriptomic changes caused by miR-7 overexpression are stronger (including miR-7-targets downregulation) and enriched in secretion/synaptic plasticity pathways. Altogether, our results suggest that in cortical neurons Cdr1as buffers miR-7 activity to control glutamatergic excitatory transmission and neuronal connectivity important for long-lasting synaptic adaptations.

SA and NHP glucosyltransferase UGT76B1 affects plant defense in both SID2- and NPR1-dependent and independent manner.

KEY MESSAGE: The small-molecule glucosyltransferase loss-of-function mutant ugt76b1 exhibits both SID2- or NPR1-dependent and independent facets of enhanced plant immunity, whereupon FMO1 is required for the SID2 and NPR1 independence. The small-molecule glucosyltransferase UGT76B1 inactivates salicylic acid (SA), isoleucic acid (ILA), and N-hydroxypipecolic acid (NHP). ugt76b1 loss-of-function plants manifest an enhanced defense status. Thus, we were interested how UGT76B1 genetically integrates in defense pathways and whether all impacts depend on SA and NHP. We study the integration of UGT76B1 by transcriptome analyses of ugt76b1. The comparison of transcripts altered by the loss of UGT76B1 with public transcriptome data reveals both SA-responsive, ISOCHORISMATE SYNTHASE 1/SALICYLIC ACID INDUCTION DEFICIENT 2 (ICS1/SID2)- and NON EXPRESSOR OF PR GENES 1 (NPR1)-dependent, consistent with the role of UGT76B1 in glucosylating SA, and SA-non-responsive, SID2/NPR1-independent genes. We also discovered that UGT76B1 impacts on a group of genes showing non-SA-responsiveness and regulation by infections independent from SID2/NPR1. Enhanced resistance of ugt76b1 against Pseudomonas syringae is partially independent from SID2 and NPR1. In contrast, the ugt76b1-activated resistance is completely dependent on FMO1 encoding the NHP-synthesizing FLAVIN-DEPENDENT MONOOXYGENASE 1). Moreover, FMO1 ranks top among the ugt76b1-induced SID2- and NPR1-independent pathogen responsive genes, suggesting that FMO1 determines the SID2- and NPR1-independent effect of ugt76b1. Furthermore, the genetic study revealed that FMO1, ENHANCED DISEASE SUSCEPTIBILITY 1 (EDS1), SID2, and NPR1 are required for the SA-JA crosstalk and senescence development of ugt76b1, indicating that EDS1 and FMO1 have a similar effect like stress-induced SA biosynthesis (SID2) or the key SA signaling regulator NPR1. Thus, UGT76B1 influences both SID2/NPR1-dependent and independent plant immunity, and the SID2/NPR1 independence is relying on FMO1 and its product NHP, another substrate of UGT76B1.

Histone Deacetylases HD2A and HD2B Undergo Feedback Regulation by ABA and Modulate Drought Tolerance via Mediating ABA-Induced Transcriptional Repression.

Histone deacetylation catalyzed by histone deacetylase plays a critical role in gene silencing and subsequently controls many important biological processes. It was reported that the expression of the plant-specific histone deacetylase subfamily HD2s is repressed by ABA in Arabidopsis. However, little is known about the molecular relationship between HD2A/HD2B and ABA during the vegetative phase. Here, we describe that the hd2ahd2b mutant shows hypersensitivity to exogenous ABA during the germination and post-germination period. Additionally, transcriptome analyses revealed that the transcription of ABA-responsive genes was reprogrammed and the global H4K5ac level is specifically up-regulated in hd2ahd2b plants. ChIP-Seq and ChIP-qPCR results further verified that both HD2A and HD2B could directly and specifically bind to certain ABA-responsive genes. As a consequence, Arabidopsis hd2ahd2b plants displayed enhanced drought resistance in comparison to WT, which is consistent with increased ROS content, reduced stomatal aperture, and up-regulated drought-resistance-related genes. Moreover, HD2A and HD2B repressed ABA biosynthesis via the deacetylation of H4K5ac at NCED9. Taken together, our results indicate that HD2A and HD2B partly function through ABA signaling and act as negative regulators during the drought resistance response via the regulation of ABA biosynthesis and response genes.

Arabidopsis histone deacetylase HD2A and HD2B regulate seed dormancy by repressing DELAY OF GERMINATION 1.

Seed dormancy is a crucial developmental transition that affects the adaption and survival of plants. Arabidopsis DELAY OF GERMINATION 1 (DOG1) is known as a master regulator of seed dormancy. However, although several upstream factors of DOG1 have been reported, the exact regulation of DOG1 is not fully understood. Histone acetylation is an important regulatory layer, controlled by histone acetyltransferases and histone deacetylases. Histone acetylation strongly correlates with transcriptionally active chromatin, whereas heterochromatin is generally characterized by hypoacetylated histones. Here we describe that loss of function of two plant-specific histone deacetylases, HD2A and HD2B, resulted in enhanced seed dormancy in Arabidopsis. Interestingly, the silencing of HD2A and HD2B caused hyperacetylation of the DOG1 locus and promoted the expression of DOG1 during seed maturation and imbibition. Knockout of DOG1 could rescue the seed dormancy and partly rescue the disturbed development phenotype of hd2ahd2b. Transcriptomic analysis of the hd2ahd2b line shows that many genes involved in seed development were impaired. Moreover, we demonstrated that HSI2 and HSL1 interact with HD2A and HD2B. In sum, these results suggest that HSI2 and HSL1 might recruit HD2A and HD2B to DOG1 to negatively regulate DOG1 expression and to reduce seed dormancy, consequently, affecting seed development during seed maturation and promoting seed germination during imbibition.

GSNOR Contributes to Demethylation and Expression of Transposable Elements and Stress-Responsive Genes.

In the past, reactive nitrogen species (RNS) were supposed to be stress-induced by-products of disturbed metabolism that cause oxidative damage to biomolecules. However, emerging evidence demonstrates a substantial role of RNS as endogenous signals in eukaryotes. In plants, S-nitrosoglutathione (GSNO) is the dominant RNS and serves as the •NO donor for S-nitrosation of diverse effector proteins. Remarkably, the endogenous GSNO level is tightly controlled by S-nitrosoglutathione reductase (GSNOR) that irreversibly inactivates the glutathione-bound NO to ammonium. Exogenous feeding of diverse RNS, including GSNO, affected chromatin accessibility and transcription of stress-related genes, but the triggering function of RNS on these regulatory processes remained elusive. Here, we show that GSNO reductase-deficient plants (gsnor1-3) accumulate S-adenosylmethionine (SAM), the principal methyl donor for methylation of DNA and histones. This SAM accumulation triggered a substantial increase in the methylation index (MI = [SAM]/[S-adenosylhomocysteine]), indicating the transmethylation activity and histone methylation status in higher eukaryotes. Indeed, a mass spectrometry-based global histone profiling approach demonstrated a significant global increase in H3K9me2, which was independently verified by immunological detection using a selective antibody. Since H3K9me2-modified regions tightly correlate with methylated DNA regions, we also determined the DNA methylation status of gsnor1-3 plants by whole-genome bisulfite sequencing. DNA methylation in the CG, CHG, and CHH contexts in gsnor1-3 was significantly enhanced compared to the wild type. We propose that GSNOR1 activity affects chromatin accessibility by controlling the transmethylation activity (MI) required for maintaining DNA methylation and the level of the repressive chromatin mark H3K9me2.

Nitric oxide coordinates growth, development, and stress response via histone modification and gene expression.

Nitric oxide (NO) is a signaling molecule with multiple regulatory functions in plant physiology and stress response. In addition to direct effects on transcriptional machinery, NO executes its signaling function via epigenetic mechanisms. We report that light intensity-dependent changes in NO correspond to changes in global histone acetylation (H3, H3K9, and H3K9/K14) in Arabidopsis (Arabidopsis thaliana) wild-type leaves, and that this relationship depends on S-nitrosoglutathione reductase and histone deacetylase 6 (HDA6). The activity of HDA6 was sensitive to NO, demonstrating that NO participates in regulation of histone acetylation. Chromatin immunoprecipitation sequencing and RNA-seq analyses revealed that NO participates in the metabolic switch from growth and development to stress response. This coordinating function of NO might be particularly important in plant ability to adapt to a changing environment, and is therefore a promising foundation for mitigating the negative effects of climate change on plant productivity.

Improving Air Quality by Nitric Oxide Consumption of Climate-Resilient Trees Suitable for Urban Greening.

Nitrogen oxides (NOx), mainly a mixture of nitric oxide (NO) and nitrogen dioxide (NO2), are formed by the reaction of nitrogen and oxygen compounds in the air as a result of combustion processes and traffic. Both deposit into leaves via stomata, which on the one hand benefits air quality and on the other hand provides an additional source of nitrogen for plants. In this study, we first determined the NO and NO2 specific deposition velocities based on projected leaf area (sV d) using a branch enclosure system. We studied four tree species that are regarded as suitable to be planted under predicted future urban climate conditions: Carpinus betulus, Fraxinus ornus, Fraxinus pennsylvanica and Ostrya carpinifolia. The NO and NO2 sVd were found similar in all tree species. Second, in order to confirm NO metabolization, we fumigated plants with 15NO and quantified the incorporation of 15N in leaf materials of these trees and four additional urban tree species (Celtis australis, Alnus spaethii, Alnus glutinosa, and Tilia henryana) under controlled environmental conditions. Based on these 15N-labeling experiments, A. glutinosa showed the most effective incorporation of 15NO. Third, we tried to elucidate the mechanism of metabolization. Therefore, we generated transgenic poplars overexpressing Arabidopsis thaliana phytoglobin 1 or 2. Phytoglobins are known to metabolize NO to nitrate in the presence of oxygen. The 15N uptake in phytoglobin-overexpressing poplars was significantly increased compared to wild-type trees, demonstrating that the NO uptake is enzymatically controlled besides stomatal dependence. In order to upscale the results and to investigate if a trade-off exists between air pollution removal and survival probability under future climate conditions, we have additionally carried out a modeling exercise of NO and NO2 deposition for the area of central Berlin. If the actually dominant deciduous tree species (Acer platanoides, Tilia cordata, Fagus sylvatica, Quercus robur) would be replaced by the species suggested for future conditions, the total annual NO and NO2 deposition in the modeled urban area would hardly change, indicating that the service of air pollution removal would not be degraded. These results may help selecting urban tree species in future greening programs.

Rapid conversion of isoprene photooxidation products in terrestrial plants.

Isoprene is emitted from the biosphere into the atmosphere, and may strengthen the defense mechanisms of plants against oxidative and thermal stress. Once in the atmosphere, isoprene is rapidly oxidized, either to isoprene-hydroxy-hydroperoxides (ISOPOOH) at low levels of nitrogen oxides, or to methyl vinyl ketone (MVK) and methacrolein at high levels. Here we combine uptake rates and deposition velocities that we obtained in laboratory experiments with observations in natural forests to show that 1,2-ISOPOOH deposits rapidly into poplar leaves. There, it is converted first to cytotoxic MVK and then most probably through alkenal/ one oxidoreductase (AOR) to less toxic methyl ethyl ketone (MEK). This detoxification process is potentially significant globally because AOR enzymes are ubiquitous in terrestrial plants. Our simulations with a global chemistry-transport model suggest that around 6.5 Tg yr- of MEK are re-emitted to the atmosphere. This is the single largest MEK source presently known, and recycles 1.5% of the original isoprene flux. Eddy covariance flux measurements of isoprene and MEK over different forest ecosystems confirm that MEK emissions can reach 1-2% those of isoprene. We suggest that detoxification processes in plants are one of the most important sources of oxidized volatile organic compounds in the atmosphere.

The Systems Architecture of Molecular Memory in Poplar after Abiotic Stress.

Throughout the temperate zones, plants face combined drought and heat spells in increasing frequency and intensity. Here, we compared periodic (intermittent, i.e., high-frequency) versus chronic (continuous, i.e., high-intensity) drought-heat stress scenarios in gray poplar (Populus× canescens) plants for phenotypic and transcriptomic effects during stress and after recovery. Photosynthetic productivity after stress recovery exceeded the performance of poplar trees without stress experience. We analyzed the molecular basis of this stress-related memory phenotype and investigated gene expression responses across five major tree compartments including organs and wood tissues. For each of these tissue samples, transcriptomic changes induced by the two stress scenarios were highly similar during the stress phase but strikingly divergent after recovery. Characteristic molecular response patterns were found across tissues but involved different genes in each tissue. Only a small fraction of genes showed similar stress and recovery expression profiles across all tissues, including type 2C protein phosphatases, the LATE EMBRYOGENESIS ABUNDANT PROTEIN4-5 genes, and homologs of the Arabidopsis (Arabidopsis thaliana) transcription factor HOMEOBOX7. Analysis of the predicted transcription factor regulatory networks for these genes suggested that a complex interplay of common and tissue-specific components contributes to the coordination of post-recovery responses to stress in woody plants.

Short-Term Exposure to Nitrogen Dioxide Provides Basal Pathogen Resistance.

Nitrogen dioxide (NO2) forms in plants under stress conditions, but little is known about its physiological functions. Here, we explored the physiological functions of NO2 in plant cells using short-term fumigation of Arabidopsis (Arabidopsis thaliana) for 1 h with 10 µL L-1 NO2. Although leaf symptoms were absent, the expression of genes related to pathogen resistance was induced. Fumigated plants developed basal disease resistance, or pattern-triggered immunity, against the necrotrophic fungus Botrytis cinerea and the hemibiotrophic bacterium Pseudomonas syringae Functional salicylic acid and jasmonic acid (JA) signaling pathways were both required for the full expression of NO2-induced resistance against B. cinerea An early peak of salicylic acid accumulation immediately after NO2 exposure was followed by a transient accumulation of oxophytodienoic acid. The simultaneous NO2-induced expression of genes involved in jasmonate biosynthesis and jasmonate catabolism resulted in the complete suppression of JA and JA-isoleucine (JA-Ile) accumulation, which was accompanied by a rise in the levels of their catabolic intermediates 12-OH-JA, 12-OH-JA-Ile, and 12-COOH-JA-Ile. NO2-treated plants emitted the volatile monoterpene α-pinene and the sesquiterpene longifolene (syn. junipene), which could function in signaling or direct defense against pathogens. NO2-triggered B. cinerea resistance was dependent on enhanced early callose deposition and CYTOCHROME P450 79B2 (CYP79B2), CYP79B3, and PHYTOALEXIN DEFICIENT3 gene functions but independent of camalexin, CYP81F2, and 4-OH-indol-3-ylmethylglucosinolate derivatives. In sum, exogenous NO2 triggers basal pathogen resistance, pointing to a possible role for endogenous NO2 in defense signaling. Additionally, this study revealed the involvement of jasmonate catabolism and volatiles in pathogen immunity.

Relationships between drought, heat and air humidity responses revealed by transcriptome-metabolome co-analysis.

BACKGROUND: Elevated temperature and reduced water availability are frequently linked abiotic stresses that may provoke distinct as well as interacting molecular responses. Based on non-targeted metabolomic and transcriptomic measurements from Arabidopsis rosettes, this study aims at a systematic elucidation of relevant components in different drought and heat scenarios as well as relationships between molecular players of stress response. RESULTS: In combined drought-heat stress, the majority of single stress responses are maintained. However, interaction effects between drought and heat can be discovered as well; these relate to protein folding, flavonoid biosynthesis and growth inhibition, which are enhanced, reduced or specifically induced in combined stress, respectively. Heat stress experiments with and without supplementation of air humidity for maintenance of vapor pressure deficit suggest that decreased relative air humidity due to elevated temperature is an important component of heat stress, specifically being responsible for hormone-related responses to water deprivation. Remarkably, this "dry air effect" is the primary trigger of the metabolomic response to heat. In contrast, the transcriptomic response has a substantial temperature component exceeding the dry air component and including up-regulation of many transcription factors and protein folding-related genes. Data level integration independent of prior knowledge on pathways and condition labels reveals shared drought and heat responses between transcriptome and metabolome, biomarker candidates and co-regulation between genes and metabolic compounds, suggesting novel players in abiotic stress response pathways. CONCLUSIONS: Drought and heat stress interact both at transcript and at metabolite response level. A comprehensive, non-targeted view of this interaction as well as non-interacting processes is important to be taken into account when improving tolerance to abiotic stresses in breeding programs. Transcriptome and metabolome may respond with different extent to individual stress components. Their contrasting behavior in response to temperature stress highlights that the protein folding machinery effectively shields the metabolism from stress. Disentangling the complex relationships between transcriptome and metabolome in response to stress is an enormous challenge. As demonstrated by case studies with supporting evidence from additional data, the large dataset provided in this study may assist in determining linked genetic and metabolic features as candidates for future mechanistic analyses.

Monoterpenes Support Systemic Acquired Resistance within and between Plants.

This study investigates the role of volatile organic compounds in systemic acquired resistance (SAR), a salicylic acid (SA)-associated, broad-spectrum immune response in systemic, healthy tissues of locally infected plants. Gas chromatography coupled to mass spectrometry analyses of SAR-related emissions of wild-type and non-SAR-signal-producing mutant plants associated SAR with monoterpene emissions. Headspace exposure of Arabidopsis thaliana to a mixture of the bicyclic monoterpenes α-pinene and ß-pinene induced defense, accumulation of reactive oxygen species, and expression of SA- and SAR-related genes, including the SAR regulatory AZELAIC ACID INDUCED1 (AZI1) gene and three of its paralogs. Pinene-induced resistance was dependent on SA biosynthesis and signaling and on AZI1 Arabidopsis geranylgeranyl reductase1 mutants with reduced monoterpene biosynthesis were SAR-defective but mounted normal local resistance and methyl salicylate-induced defense responses, suggesting that monoterpenes act in parallel with SA The volatile emissions from SAR signal-emitting plants induced defense in neighboring plants, and this was associated with the presence of α-pinene, ß-pinene, and camphene in the emissions of the "sender" plants. Our data suggest that monoterpenes, particularly pinenes, promote SAR, acting through ROS and AZI1, and likely function as infochemicals in plant-to-plant signaling, thus allowing defense signal propagation between neighboring plants.

Nitric Oxide Modulates Histone Acetylation at Stress Genes by Inhibition of Histone Deacetylases.

Histone acetylation, which is an important mechanism to regulate gene expression, is controlled by the opposing action of histone acetyltransferases and histone deacetylases (HDACs). In animals, several HDACs are subjected to regulation by nitric oxide (NO); in plants, however, it is unknown whether NO affects histone acetylation. We found that treatment with the physiological NO donor S-nitrosoglutathione (GSNO) increased the abundance of several histone acetylation marks in Arabidopsis (Arabidopsis thaliana), which was strongly diminished in the presence of the NO scavenger 2-4-carboxyphenyl-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide. This increase was likely triggered by NO-dependent inhibition of HDAC activity, since GSNO and S-nitroso-N-acetyl-dl-penicillamine significantly and reversibly reduced total HDAC activity in vitro (in nuclear extracts) and in vivo (in protoplasts). Next, genome-wide H3K9/14ac profiles in Arabidopsis seedlings were generated by chromatin immunoprecipitation sequencing, and changes induced by GSNO, GSNO/2-4-carboxyphenyl-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide or trichostatin A (an HDAC inhibitor) were quantified, thereby identifying genes that display putative NO-regulated histone acetylation. Functional classification of these genes revealed that many of them are involved in the plant defense response and the abiotic stress response. Furthermore, salicylic acid, which is the major plant defense hormone against biotrophic pathogens, inhibited HDAC activity and increased histone acetylation by inducing endogenous NO production. These data suggest that NO affects histone acetylation by targeting and inhibiting HDAC complexes, resulting in the hyperacetylation of specific genes. This mechanism might operate in the plant stress response by facilitating the stress-induced transcription of genes.

Nitrite is the driver, phytohormones are modulators while NO and H2O2 act as promoters of NO2-induced cell death.

This study aimed to understand the molecular mechanisms of nitrogen dioxide (NO2)-induced toxicity and cell death in plants. Exposure of Arabidopsis to high concentrations of NO2 induced cell death in a dose-dependent manner. No leaf symptoms were visible after fumigation for 1 h with 10 parts per million (ppm) NO2 However, 20 ppm NO2 caused necrotic lesion formation and 30 ppm NO2 complete leaf collapse, which had already started during the 1 h fumigation period. NO2 fumigation resulted in a massive accumulation of nitrite and in protein modifications by S-nitrosylation and tyrosine nitration. Nitric oxide (NO) at 30 ppm did not trigger leaf damage or any of the effects observed after NO2 fumigation. The onset of NO2-induced cell death correlated with NO and hydrogen peroxide (H2O2) signaling and a decrease in antioxidants. NO- and H2O2-accumulating mutants were more sensitive to NO2 than wild-type plants. Accordingly, experiments with specific scavengers confirmed that NO and H2O2 are essential promoters of NO2-induced cell death. Leaf injection of 100 mM nitrite caused an increase in S-nitrosylation, NO, H2O2, and cell death suggesting that nitrite functioned as a mediator of NO2-induced effects. A targeted screening of phytohormone mutants revealed a protective role of salicylic acid (SA) signaling in response to NO2 It was also shown that phytohormones were modulators rather than inducers of NO2-induced cell death. The established experimental set-up is a suitable system to investigate NO2 and cell death signaling in large-scale mutant screens.

Toward computational cumulative biology by combining models of biological datasets.

A main challenge of data-driven sciences is how to make maximal use of the progressively expanding databases of experimental datasets in order to keep research cumulative. We introduce the idea of a modeling-based dataset retrieval engine designed for relating a researcher's experimental dataset to earlier work in the field. The search is (i) data-driven to enable new findings, going beyond the state of the art of keyword searches in annotations, (ii) modeling-driven, to include both biological knowledge and insights learned from data, and (iii) scalable, as it is accomplished without building one unified grand model of all data. Assuming each dataset has been modeled beforehand, by the researchers or automatically by database managers, we apply a rapidly computable and optimizable combination model to decompose a new dataset into contributions from earlier relevant models. By using the data-driven decomposition, we identify a network of interrelated datasets from a large annotated human gene expression atlas. While tissue type and disease were major driving forces for determining relevant datasets, the found relationships were richer, and the model-based search was more accurate than the keyword search; moreover, it recovered biologically meaningful relationships that are not straightforwardly visible from annotations-for instance, between cells in different developmental stages such as thymocytes and T-cells. Data-driven links and citations matched to a large extent; the data-driven links even uncovered corrections to the publication data, as two of the most linked datasets were not highly cited and turned out to have wrong publication entries in the database.

Integrative and personalized QSAR analysis in cancer by kernelized Bayesian matrix factorization.

With data from recent large-scale drug sensitivity measurement campaigns, it is now possible to build and test models predicting responses for more than one hundred anticancer drugs against several hundreds of human cancer cell lines. Traditional quantitative structure-activity relationship (QSAR) approaches focus on small molecules in searching for their structural properties predictive of the biological activity in a single cell line or a single tissue type. We extend this line of research in two directions: (1) an integrative QSAR approach predicting the responses to new drugs for a panel of multiple known cancer cell lines simultaneously and (2) a personalized QSAR approach predicting the responses to new drugs for new cancer cell lines. To solve the modeling task, we apply a novel kernelized Bayesian matrix factorization method. For maximum applicability and predictive performance, the method optionally utilizes genomic features of cell lines and target information on drugs in addition to chemical drug descriptors. In a case study with 116 anticancer drugs and 650 cell lines, we demonstrate the usefulness of the method in several relevant prediction scenarios, differing in the amount of available information, and analyze the importance of various types of drug features for the response prediction. Furthermore, after predicting the missing values of the data set, a complete global map of drug response is explored to assess treatment potential and treatment range of therapeutically interesting anticancer drugs.

A community effort to assess and improve drug sensitivity prediction algorithms.

Predicting the best treatment strategy from genomic information is a core goal of precision medicine. Here we focus on predicting drug response based on a cohort of genomic, epigenomic and proteomic profiling data sets measured in human breast cancer cell lines. Through a collaborative effort between the National Cancer Institute (NCI) and the Dialogue on Reverse Engineering Assessment and Methods (DREAM) project, we analyzed a total of 44 drug sensitivity prediction algorithms. The top-performing approaches modeled nonlinear relationships and incorporated biological pathway information. We found that gene expression microarrays consistently provided the best predictive power of the individual profiling data sets; however, performance was increased by including multiple, independent data sets. We discuss the innovations underlying the top-performing methodology, Bayesian multitask MKL, and we provide detailed descriptions of all methods. This study establishes benchmarks for drug sensitivity prediction and identifies approaches that can be leveraged for the development of new methods.

Dense module enumeration in biological networks.

Automatic discovery of functional complexes from protein interaction data is a rewarding but challenging problem. While previous approaches use approximations to extract dense modules, our approach exactly solves the problem of dense module enumeration. Furthermore, constraints from additional information sources such as gene expression and phenotype data can be integrated, so we can systematically detect dense modules with interesting profiles. Given a weighted protein interaction network, our method discovers all protein sets that satisfy a user-defined minimum density threshold. We employ a reverse search strategy, which allows us to exploit the density criterion in an efficient way.

Targeted retrieval of gene expression measurements using regulatory models.

MOTIVATION: Large public repositories of gene expression measurements offer the opportunity to position a new experiment into the context of earlier studies. While previous methods rely on experimental annotation or global similarity of expression profiles across genes or gene sets, we compare experiments by measuring similarity based on an unsupervised, data-driven regulatory model around pre-specified genes of interest. Our experiment retrieval approach is novel in two conceptual respects: (i) targetable focus and interpretability: the analysis is targeted at regulatory relationships of genes that are relevant to the analyst or come from prior knowledge; (ii) regulatory model-based similarity measure: related experiments are retrieved based on the strength of inferred regulatory links between genes. RESULTS: We learn a model for the regulation of specific genes from a data repository and exploit it to construct a similarity metric for an information retrieval task. We use the Fisher kernel, a rigorous similarity measure that typically has been applied to use generative models in discriminative classifiers. Results on human and plant microarray collections indicate that our method is able to substantially improve the retrieval of related experiments against standard methods. Furthermore, it allows the user to interpret biological conditions in terms of changes in link activity patterns. Our study of the osmotic stress network for Arabidopsis thaliana shows that the method successfully identifies relevant relationships around given key genes. AVAILABILITY: The code (R) is available at http://research.ics.tkk.fi/mi/software.shtml. CONTACT: elisabeth.georgii@aalto.fi; jarkko.salojarvi@helsinki.fi; samuel.kaski@hiit.fi SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Enumeration of condition-dependent dense modules in protein interaction networks.

MOTIVATION: Modern systems biology aims at understanding how the different molecular components of a biological cell interact. Often, cellular functions are performed by complexes consisting of many different proteins. The composition of these complexes may change according to the cellular environment, and one protein may be involved in several different processes. The automatic discovery of functional complexes from protein interaction data is challenging. While previous approaches use approximations to extract dense modules, our approach exactly solves the problem of dense module enumeration. Furthermore, constraints from additional information sources such as gene expression and phenotype data can be integrated, so we can systematically mine for dense modules with interesting profiles. RESULTS: Given a weighted protein interaction network, our method discovers all protein sets that satisfy a user-defined minimum density threshold. We employ a reverse search strategy, which allows us to exploit the density criterion in an efficient way. Our experiments show that the novel approach is feasible and produces biologically meaningful results. In comparative validation studies using yeast data, the method achieved the best overall prediction performance with respect to confirmed complexes. Moreover, by enhancing the yeast network with phenotypic and phylogenetic profiles and the human network with tissue-specific expression data, we identified condition-dependent complex variants. AVAILABILITY: A C++ implementation of the algorithm is available at http://www.kyb.tuebingen.mpg.de/~georgii/dme.html. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.