Enrichment analysis for discovering biological associations in phenotypic screens.

A phenotypic screen (PS) is used to identify compounds causing a desired phenotype in a complex biological system where mechanisms and targets are largely unknown. Deconvoluting the mechanism of action of actives and identification of relevant targets and pathways remains a formidable challenge. Current methods fail to use the rich information available regarding compounds and their targets in a systematic way for this deconvolution. We have developed an enrichment analysis algorithm to identify targets associated with the desired phenotype in a rigorous data-driven manner using actives and hundreds of thousands of inactives in a PS, as well as results of thousands of available legacy target-based screens in an institution. Our method quantifies association between the PS and targets while reducing sampling bias, which leads to identification of novel targets, additional chemical matter, and appropriate assays. Its use is illustrated using two examples from our laboratories: TRAIL and DNA fragmentation. Enrichment analysis of these PSs is discussed using both biological pathway analysis and known cell biology to demonstrate the value of our method. We believe this enrichment analysis method is an indispensable tool for the analysis of PSs.

OHMM: a Hidden Markov Model accurately predicting the occupancy of a transcription factor with a self-overlapping binding motif.

BACKGROUND: DNA sequence binding motifs for several important transcription factors happen to be self-overlapping. Many of the current regulatory site identification methods do not explicitly take into account the overlapping sites. Moreover, most methods use arbitrary thresholds and fail to provide a biophysical interpretation of statistical quantities. In addition, commonly used approaches do not include the location of a site with respect to the transcription start site (TSS) in an integrated probabilistic framework while identifying sites. Ignoring these features can lead to inaccurate predictions as well as incorrect design and interpretation of experimental results. RESULTS: We have developed a tool based on a Hidden Markov Model (HMM) that identifies binding location of transcription factors with preference for self-overlapping DNA motifs by combining the effects of their alternative binding modes. Interpreting HMM parameters as biophysical quantities, this method uses the occupancy probability of a transcription factor on a DNA sequence as the discriminant function, earning the algorithm the name OHMM: Occupancy via Hidden Markov Model. OHMM learns the classification threshold by training emission probabilities using unaligned sequences containing known sites and estimating transition probabilities to reflect site density in all promoters in a genome. While identifying sites, it adjusts parameters to model site density changing with the distance from the transcription start site. Moreover, it provides guidance for designing padding sequences in gel shift experiments. In the context of binding sites to transcription factor NF-kappaB, we find that the occupancy probability predicted by OHMM correlates well with the binding affinity in gel shift experiments. High evolutionary conservation scores and enrichment in experimentally verified regulated genes suggest that NF-kappaB binding sites predicted by our method are likely to be functional. CONCLUSION: Our method deals specifically with identifying locations with multiple overlapping binding sites by computing the local occupancy of the transcription factor. Moreover, considering OHMM as a biophysical model allows us to learn the classification threshold in a principled manner. Another feature of OHMM is that we allow transition probabilities to change with location relative to the TSS. OHMM could be used to predict physical occupancy, and provides guidance for proper design of gel-shift experiments. Based upon our predictions, new insights into NF-kappaB function and regulation and possible new biological roles of NF-kappaB were uncovered.

Molecular characterization of antigen-induced lung inflammation in a murine model of asthma.

Asthma is one of the foremost contributors to morbidity and mortality in industrialized countries. Our objective was to characterize the acute response to allergen and to identify potentially novel molecular targets for pharmacological intervention in asthma. We therefore designed a study to identify genes whose regulation was altered following ovalbumin (OVA) challenge in the presence and absence of treatment with glucocorticoids in BALB/c mice. RNA was isolated from lungs for gene profiling from 8-week-old sensitized mice, 3 and 18 hours post OVA challenge on days 1, 4, and 7 of aerosol challenge. Taqman (real time RT-PCR) analysis of marker genes indicative of Th2 (IL-4, IL-13), eosinophil (RANTES, eotaxin), Th1/macrophage (IFNgamma) and epithelial cell (MUC5AC) phenotypes were used to characterize responses to allergen challenge. Histological evaluation of lungs from additional challenged animals revealed inflammatory infiltrates on days 4 and 7, but not on day 1 post challenge. We postulate that expression of IL-4, IL-13 and other genes by OVA at day 1 probably reflects activation of resident cells, whereas the fivefold increase in the number of regulated genes at day 7 reflects the contribution of recruited cells. Of the regulated genes, only a subset was counter-regulated by dexamethasone treatment. Although regulated genes included genes in many protein families, herein we report regulation of two proteases whose role in response to OVA challenge has not been characterized. This model will be used to generate disease hypotheses for which may play an important role in initiating disease pathology in this model.

Repression of B-cell linker (BLNK) and B-cell adaptor for phosphoinositide 3-kinase (BCAP) is important for lymphocyte transformation by rel proteins.

Persistent Rel/nuclear factor-kappaB (NF-kappaB) activity is a hallmark of many human cancers, and the Rel proteins are implicated in leukemia/lymphomagenesis but the mechanism is not fully understood. Microarray analysis to identify transformation-impacting genes regulated by NF-kappaB's oncogenic v-Rel and c-Rel proteins uncovered that Rel protein expression leads to transcriptional repression of key B-cell receptor (BCR) components and signaling molecules like B-cell linker (BLNK), the B-cell adaptor for phosphoinositide 3-kinase (BCAP) and immunoglobulin lambda light chain (Ig lambda), and is accompanied by a block in BCR-mediated activation of extracellular signal-regulated kinase, Akt, and c-Jun-NH(2)-kinase in response to anti-IgM. The BLNK and BCAP proteins were also down-regulated in lymphoid cells expressing a transformation-competent chimeric RelA/v-Rel protein, suggesting a correlation with the capacity of Rel proteins to transform lymphocytes. DNA-binding studies identified functional NF-kappaB-binding sites, and chromatin immunoprecipitation (ChIP) data showed binding of Rel to the endogenous blnk and bcap promoters in vivo. Importantly, restoration of either BLNK or BCAP expression strongly inhibited transformation of primary chicken lymphocytes by the potent NF-kappaB oncoprotein v-Rel. These findings are interesting because blnk and other BCR components and signaling molecules are down-regulated in primary mediastinal large B-cell lymphomas and Hodgkin's lymphomas, which depend on c-Rel for survival, and are consistent with the tumor suppressor function of BLNK. Overall, our results indicate that down-regulation of BLNK and BCAP is an important contributing factor to the malignant transformation of lymphocytes by Rel and suggest that gene repression may be as important as transcriptional activation for Rel's transforming activity.

Subcellular localization of the yeast proteome.

Protein localization data are a valuable information resource helpful in elucidating eukaryotic protein function. Here, we report the first proteome-scale analysis of protein localization within any eukaryote. Using directed topoisomerase I-mediated cloning strategies and genome-wide transposon mutagenesis, we have epitope-tagged 60% of the Saccharomyces cerevisiae proteome. By high-throughput immunolocalization of tagged gene products, we have determined the subcellular localization of 2744 yeast proteins. Extrapolating these data through a computational algorithm employing Bayesian formalism, we define the yeast localizome (the subcellular distribution of all 6100 yeast proteins). We estimate the yeast proteome to encompass approximately 5100 soluble proteins and >1000 transmembrane proteins. Our results indicate that 47% of yeast proteins are cytoplasmic, 13% mitochondrial, 13% exocytic (including proteins of the endoplasmic reticulum and secretory vesicles), and 27% nuclear/nucleolar. A subset of nuclear proteins was further analyzed by immunolocalization using surface-spread preparations of meiotic chromosomes. Of these proteins, 38% were found associated with chromosomal DNA. As determined from phenotypic analyses of nuclear proteins, 34% are essential for spore viability--a percentage nearly twice as great as that observed for the proteome as a whole. In total, this study presents experimentally derived localization data for 955 proteins of previously unknown function: nearly half of all functionally uncharacterized proteins in yeast. To facilitate access to these data, we provide a searchable database featuring 2900 fluorescent micrographs at http://ygac.med.yale.edu.