Genomic privacy and limits of individual detection in a pool
Room: Ballroom A/B
Presenting author: Sriram Sankararaman, University of California, Berkeley, United States

Additional authors:
Guillaume Obozinski, INRIA , France
Michael Jordan, UC Berkeley, United States
Eran Halperin, ICSI, Tel Aviv University, Israel

Presentation Overview:
Methods for the detection of a single individual in summary data from genome-wide association studies have recently been shown to have sufficient power to jeopardize the privacy of the study's subjects. We present an analytical and empirical study of the statistical power of such methods. The analysis aims to provide quantitative guidelines for researchers wishing to make a limited number of SNPs available publicly without compromising privacy.

TOP

Species Tree Inference by Minimizing Deep Coalescences
Room: Ballroom C
Presenting author: Luay Nakhleh, Rice University, United States

Additional authors:
Cuong Than, University of Michigan, United States

Presentation Overview:
The traditional approach to species tree inference entails (1) sequencing a single locus from the group of species under study, (2) reconstructing a "gene tree" for this locus, and (3) declaring this tree to be the species tree. However, the increasing availability of whole-genome and multi-locus data from multiple organisms and populations has highlighted the issue of species/gene tree incongruence and, consequently, the inappropriateness of the traditional approach...In this talk, I will present the issue of incomplete lineage sorting, its effects on gene/species tree congruence, and our recent approaches for inferring species trees despite lineage sorting. Using these approaches, we analyze biological, as well as synthetic, data sets, and demonstrate the accuracy of the species tree estimates obtained. Our method is very fast, which makes it appropriate for analyzing data at a genomic scale.

TOP

Can literature analysis identify innovation drivers in drug discovery?
Room: Ballroom A/B
Presenting author: Pankaj Agarwal, GlaxoSmithKline, United States


Presentation Overview:
Bioinformatics methodology plays multiple roles in the pharmaceutical industry ranging from target identification and identifying disease indications for drugs to guiding strategic decisions. This talk will cover aspects of this spectrum, but it will primarily focus on identifying novel areas of science. Drug discovery must be guided not only by medical need and commercial potential, but also by the areas in which new science is creating therapeutic opportunities, such as target identification and the understanding of disease mechanisms. To systematically identify such areas of high scientific activity, we use bibliometrics and related data-mining methods to analyse over half a terabyte of data, including PubMed abstracts, literature citation data and patent filings. These analyses reveal trends in scientific activity related to disease studied at varying levels, down to individual genes and pathways, and provide methods to monitor areas in which scientific advances are likely to create new therapeutic opportunities.

TOP

DNA Sudoku
Room: Ballroom C
Presenting author: Or Zuk, Broad Institute, United States

Additional authors:
Assaf Gordon, Cold Spring Harbor, United States
Ken Chang, Cold Spring Harbor, United States
Greg Hannon, Cold Spring Harbor, United States
Roy Ronen, Tel-Aviv University, Israel
Partha Mitra, Cold Spring Harbor, United States
Oron Navon, Tel-Aviv University, Israel
Michelle Rooks, Cold Spring Harbor, United States
Michael Brand, Lester Associates, Australia

Presentation Overview:
We devised a novel sequencing strategy, called "DNA Sudoku", which enables multiplexing thousands of specimens in a single run of next-gen sequencing. Current multiplexing schemes are based on ligating DNA barcodes to each specimen. While this approach is adequate for small-scale experiments, it poorly scales, as massive barcode synthesis is laborious and expensive. DNA Sudoku uses combinatorial pooling, which drastically reduces the number of barcodes. It is reminiscent of solving a Sudoku puzzle: every specimen is like a cell in the puzzle, and the genotypes are the digits. After solving the genotype of one specimen, we propagate the information to the other specimens in the pool, and solve their genotypes. We repeat that process until all genotypes are solved. DNA Sudoku has important implications in medical genetics, particularly for carrier screens. We offer a method to harness the power of next-gen sequencers to develop a comprehensive and cost-effective genetic screen.

TOP

Finding co-occurrence of copy number changes in tumors.
Room: Ballroom A/B
Presenting author: Christiaan Klijn, Netherlands Cancer Institute, Netherlands

Additional authors:
Jan Bot, Delft University of Technology, Netherlands
Marcel Reinders, Delft University of Technology, Netherlands
David Adams, The Wellcome Trust Sanger Institute, United Kingdom
Lodewyk Wessels, Netherlands Cancer Institute, Netherlands
Jos Jonkers, Netherlands Cancer Institute, Netherlands

Presentation Overview:
Cancer, a genetic disease, is only rarely caused by a single mutation. Often, multiple (epi)genetic mutations have to occur to cause oncogenic transformation. Cancer genes can be disrupted by many different processes of which DNA copy number change due to genomic instability is one. Although many studies have addressed the detection of important single copy number changes in different tumor types, no studies have thoroughly investigated the interdependence of copy number changes Our study focuses on the detection of co-occurrent DNA copy number changes. We show that, in lymphoid tumors, there are large networks of related copy number changes over the entire genome. The genomic locations that show significant co-occurrent copy number changes are significantly enriched for functionally related and cancer associated genes. Our observations support the notion that subtle gene dosage changes of many related genes could be an overlooked but important process in cancer.

TOP

Genome-wide Identification of Human RNA Editing Sites by Massively Parallel DNA Capturing and Sequencing
Room: Ballroom C
Presenting author: Erez Levanon, Bar-Ilan University, Israel

Additional authors:
Jin Billy Li, Harvard Medical School, United States
Jung-Ki Yoon, Seoul National University, South Korea
John Aach, Harvard Medical School, United States
Bin Xie, Virginia Commonwealth University, United States
Emily LeProust, Agilent Technologies, United States
Yuan Gao, Virginia Commonwealth University, United States
Kun Zhang, University of California, San Diego, United States
George Church, Harvard Medical School, United States

Presentation Overview:
Adenosine-to-inosine (A-to-I) RNA editing leads to transcriptome diversity and is important for normal brain function. To date, only a handful of functional sites have been identified in mammals. We developed an unbiased assay to screen more than 36,000 computationally predicted nonrepetitive A-to-I sites using massively parallel target capture and DNA sequencing. A comprehensive set of several hundred human RNA editing sites was detected by comparing genomic DNA with RNAs from seven tissues of a single individual. Specificity of our profiling was supported by observations of enrichment with known features of targets of adenosine deaminases acting on RNA (ADAR) and validation by means of capillary sequencing. This efficient approach greatly expands the repertoire of RNA editing targets and can be applied to studies involving RNA editingÐrelated human diseases.

TOP

Novel statistics reveal cancer universal miRNA activity
Room: Ballroom A/B
Presenting author: Roy Navon, Agilent Laboratories, Israel

Additional authors:
Hui Wang, Agilent Laboratories, United States
Israel Steinfeld, Technion, Israel
Anya Tsalenko, Agilent Laboratories, United States
Amir Ben-Dor, Agilent Laboratories, Israel
Zohar Yakhini, Agilent Laboratories, Israel

Presentation Overview:
microRNAs (miRNAs) regulate genes and play important roles in cancer pathogenesis and development. Variation amongst individuals is a significant confounding factor in miRNA (or other) expression studies. The true character of biologically or clinically meaningful differential expression can be obscured by inter-patient variation. We will present data from microarray profiling of more than 700 miRNAs in 28 matched (same patient) tumor/normal samples from 8 different tumor types (breast, colon, liver, lung, lymphoma, ovary, prostate and testis) - a design that minimizes tissue type and patient related variability. We will then describe novel statistical methods used in analyzing this data. The analysis revealed several miRNA that are consistently differentially expressed over multiple tumor types. These differentially expressed miRNAs include known oncomiRs as well as miRNAs that were not previously universally associated with cancer, such as miR-133b and miR-486-5p, both consistently down regulated in cancer, in the context of our cohort.

TOP

The effect of histone sequence preferences on nucleosome organization and gene regulation
Room: Ballroom C
Presenting author: Noam Kaplan, The Weizmann Institute, Israel

Additional authors:
Irene K Moore, Northwestern University, United States
Yvonne Fondufe-Mittendorf, Northwestern University, United States
Andrea J Gossett, University of North Carolina at Chapel Hill, United States
Desiree Tillo, University of Toronto, Canada
Yair Field, The Weizmann Institute of Science, Israel
Emily M LeProust, Agilent Technologies, United States
Timothy R Hughes, University of Toronto, Canada
Jason D Lieb, University of North Carolina at Chapel Hill, United States
Jon Widom, Northwestern University, United States
Eran Segal, The Weizmann Institute of Science, Israel

Presentation Overview:
Nucleosome organization is critical for gene regulation. In vivo, nucleosome organization is determined by multiple factors including transcription factors, chromatin remodellers and DNA sequence preferences of the nucleosomes themselves. Here we determine the importance of histone sequence preferences experimentally by measuring the genome-wide occupancy of nucleosomes assembled on purified yeast DNA. The resulting map, in which nucleosomes are governed only by their sequence preferences, is similar to in vivo nucleosome maps generated in three different growth conditions. In vitro, nucleosome depletion is evident at transcription factor binding sites and around gene start and end sites, indicating that nucleosome depletion at these sites in vivo is partly encoded in the genome. Using our in vitro data, we devise a computational model of nucleosome sequence preferences that is predictive of in vivo occupancy in worm. Our results indicate that histone sequence preferences have a central role in determining nucleosome organization in vivo.

TOP

High-throughput analysis of miRNAs regulating the Estrogen Receptor
Room: Ballroom A/B
Presenting author: Pekka Kohonen, University of Turku, Finland

Additional authors:
Suvi-Katri Leivonen, VTT Technical Research Centre of Finland, and Centre for Biotechnology, University of Turku, Finland
Rami Mäkela, VTT Technical Research Centre of Finland, and Centre for Biotechnology, University of Turku, Finland
Päivi Östling, VTT Technical Research Centre of Finland, and Centre for Biotechnology, University of Turku, Finland
Saija Haapa-Paananen, VTT Technical Research Centre of Finland, and Centre for Biotechnology, University of Turku, Finland
Kristine Kleivi, Institute for Cancer Research, Rikshospitalet-Radiumhospitalet Medical Center, Norway
Espen Enerly, Institute for Cancer Research, The Norwegian Radium Hospital, Norway
Anna Aakula, VTT Technical Research Centre of Finland, and Centre for Biotechnology, University of Turku, Finland
Niko Sahlberg, Biotechnology Centre of Oslo, Norway
Vessela Kristensen, Faculty of Medicine, University of Oslo, Norway
Anne-Lise Børresen-Dale, Institute for Cancer Research Norwegian Radium Hospital Rikshospitalet University Hospital, Norway
Kirsi Hellstršm, University of Helsinki, Finland
Petri Saviranta, VTT Technical Research Centre of Finland, and University of Turku, Finland
Merja Perälä, VTT Technical Research Centre of Finland, and University of Turku, Finland
Olli Kallioniemi, University of Helsinki and VTT Technical Research Centre of Finland, Finland

Presentation Overview:
Predicting the impact of miRNAs on target proteins is challenging because of their different regulatory effects at the transcriptional and especially translational levels. In this study, we demonstrate the high-throughput protein lysate microarray (LMA) technology as a novel, powerful technique in determining the relative impact of various miRNAs on key target proteins and associated cellular processes and pathways. Target protein levels of 319 pre-miRs were monitored after high-throughput transfections into breast cancer cells. We identified and validated 21 miRNAs that downregulated the estrogen receptor-alpha (ERalpha). Five potent ERalpha-regulating miRNAs were confirmed to directly target ERalpha in 3'-UTR reporter assays. The gene expression signature that they repressed highly overlapped with that of a siRNA against ERalpha, and across all the signatures tested, was most closely associated with the repression of known estrogen-induced genes. Furthermore, miR-18a and miR-18b showed higher levels of expression in ERalpha-negative as compared with ERalpha-positive clinical tumors.

TOP

Cracking the genetic code of heart regulatory elements
Room: Ballroom C
Presenting author: Ivan Ovcharenko, National Institutes of Health (NIH), United States

Additional authors:
Leelavati Narlikar, NIH, United States
Noboru Sakabe, University of Chicago, United States
Alexander Blanski, University of Chicago, United States
Fabio Arimura, University of Chicago, United States
John Westlund, University of Chicago, United States
Marcelo Nobrega, University of Chicago, United States

Presentation Overview:
We describe a computational strategy to systematically identify heart-specific cis-regulatory elements. We employed a combination of Gibbs sampling and linear regression to build a classifier that identifies heart enhancers based on the pattern of sequence features, including known and putative transcription factor (TF) binding specificities. In detecting heart enhancers, the 92% accuracy of our classifier is vastly superior to other commonly used methods. Furthermore, most of the predicted binding specificities resemble the specificities of TFs active in heart development, such as SRF, MEF2, SMAD, and GATA. Using our classifier as a predictor, a genome-wide scan identified over 40,000 novel human heart enhancers that are strongly associated with genes expressed in the heart. Finally, in vivo tests of our predictions in mouse and zebrafish achieved a validation rate of 62%, significantly higher than what is expected by chance. These results support the existence of computationally amenable cis-regulatory encryption in mammalian genomes.

TOP

The robustness of disease signatures across tissues and experiments
Room: Ballroom A/B
Presenting author: Joel Dudley, Stanford University School of Medicine, United States

Additional authors:
Robert Tibshirani, Stanford University, United States
Tarangini Deshpande, NuMedii Inc., United States
Atul Butte, Stanford University, United States

Presentation Overview:
Meta-analyses combining gene expression microarray experiments offer new insights into the molecular pathophysiology of disease not evident from individual experiments. However, the pathophysiological reproducibility across experiments is not well established. In this study, we carried out a large-scale analysis of experiments obtained from NCBI GEO, and evaluated their concordance across a broad range of diseases and tissue types. Evaluating 238 diseases and 122 tissues from 8435 microarrays, we find evidence for a pathophysiological concordance between experiments measuring the same disease condition. Furthermore, we ?nd that the molecular signature of disease across tissues is overall more prominent than the signature of tissue expression across diseases. The results offer new insight into the quality of public microarray data using pathophysiological metrics, and support new research directions into the commonalities of disease irrespective of tissue, as well as the creation of multi-tissue systems models of disease pathology using public data.

TOP

Interrogating high-level DNA structure from ChIA-PET Data
Room: Ballroom C
Presenting author: Han Xu, Genome Institute of Singapore, Singapore

Additional authors:
Melissa J. Fullwood, Genome Institute of Singapore, Singapore
Guoliang Li, Genome Institute of Singapore, Singapore
Fabianus Hendriyan Mulawadi, Genome Institute of Singapore, Singapore
Stoyan Velkov, Genome Institute of Singapore, Singapore
Vinsensius B. Vega, Genome Institute of Singapore, Singapore
Pramila N. Ariyaratne, Genome Institute of Singapore, Singapore
Yusoff Bin Mohamed, Genome Institute of Singapore, Singapore
Hong Sain Ooi, Genome Institute of Singapore, Singapore
Chandana Tennakoon, National University of Singapore, Singapore
Mei Hui Liu, Genome Institute of Singapore, Singapore
You Fu Pan, Genome Institute of Singapore, Singapore
Yuriy L. Orlov, Genome Institute of Singapore, Singapore
Andrea Ho, Genome Institute of Singapore, Singapore
Poh Huay Mei, Genome Institute of Singapore, Singapore
Elaine G. Y. Chew, Genome Institute of Singapore, Singapore
Phillips Yao Hui Huang, Genome Institute of Singapore, Singapore
Jun Liu, Genome Institute of Singapore, Singapore
Willem-Jan Welboren, Radboud University, Netherlands
Yuyuan Han, Genome Institute of Singapore, Singapore
Yanquan Luo, Genome Institute of Singapore, Singapore
Peck Yean Tan, Genome Institute of Singapore, Singapore
Pei Ye Choy, Genome Institute of Singapore, Singapore
K. D. Senali Abayratna Wansa, Genome Institute of Singapore, Singapore
Bing Zhao, Genome Institute of Singapore, Singapore
Kar Sian Lim, Genome Institute of Singapore, Singapore
Shi Chi Leow, Genome Institute of Singapore, Singapore
Jit Sin Yow, Genome Institute of Singapore, Singapore
Roy Joseph, Genome Institute of Singapore, Singapore
Haixia Li, Genome Institute of Singapore, Singapore
Kartiki V. Desai, Genome Institute of Singapore, Singapore
Jane S. Thomsen, Genome Institute of Singapore, Singapore
Yew Kok Lee, Genome Institute of Singapore, Singapore
R. Krishna Murthy Karuturi, Genome Institute of Singapore, Singapore
Thoreau Herve, Genome Institute of Singapore, Singapore
Guillaume Bourque, Genome Institute of Singapore, Singapore
Hendrik G. Stunnenberg, Radboud University, Netherlands
Xiaoan Ruan, Genome Institute of Singapore, Singapore
Valere Cacheux-Rataboul, Genome Institute of Singapore, Singapore
Edison T. Liu, Genome Institute of Singapore, Singapore
Chia-Lin Wei, Genome Institute of Singapore, Singapore
Edwin Cheung, Genome Institute of Singapore, Singapore
Yijun Ruan, Genome Institute of Singapore, Singapore
Wing-Kin Sung, Genome Institute of Singapore, Singapore

Presentation Overview:
Although the genome is linear, it is organized as a complex three-dimensional structure, through which DNA elements separated by long genomic distances can in principle interact with each other. We developed a new strategy, chromatin interaction analysis by paired-end-tag sequencing (ChIA-PET), for the de novo detection of the global chromatin interactions. Analysis of the ChIA-PET dataset posed new challenges to the area of bioinformatics and computational biology. In this talk, we will present: a) a computational pipeline for identifying chromatin interactions from ChIA-PET data; b) highlighted results from an ER-alpha dataset; c) structural models of chromatin interactions implied by ChIA-PET observations; d) open problems related to the computational analysis of ChIA-PET.

TOP

Predicting Positive p53 Cancer Rescue Regions Using Most Informative Positive (MIP) Active Learning
Room: Ballroom A/B
Presenting author: Richard Lathrop, University of California, Irvine, United States

Additional authors:
Samuel Danziger, Univ. of California, Irvine, United States
Roberta Baronio, Univ. of California, Irvine, United States
Lydia Ho, Univ. of California, Irvine, United States
Linda Hall, Univ. of California, Irvine, United States
Kirsty Salmon, Verdezyne, Inc., United States
Wesley Hatfield, Univ. of California, Irvine, United States
Peter Kaiser, Univ. of California, Irvine, United States

Presentation Overview:
Many protein engineering problems involve finding mutations that produce proteins with a particular function. Most Informative Positive (MIP) active learning is tailored to biological problems because it seeks novel and informative positive results. We applied MIP to discover mutations in the tumor suppressor protein p53 that reactivate mutated p53 found in human cancers. MIP found Positive (cancer rescue) p53 mutants in silico using 33% fewer experiments than traditional non-MIP active learning. MIP was used to select a Positive Region predicted to be enriched for p53 cancer rescue mutants. In vivo assays showed that the predicted Positive Region: (1) had significantly more (p<0.01) new strong cancer rescue mutants than control regions (Negative, and non-MIP active learning); (2) had slightly more new strong cancer rescue mutants than an Expert region selected for purely biological considerations; and (3) rescued for the first time the previously unrescuable p53 cancer mutant P152L.

TOP

Systems-level dynamic analyses of fate change in murine embryonic stem cells
Room: Ballroom C
Presenting author: Florian Markowetz, Cancer Research UK, United Kingdom

Additional authors:
Anthony Whetton, Uni Manchester, United Kingdom
Richard Unwin, Uni Manchester, United Kingdom
Edoardo Airoldi, Harvard, United States
Laurie Boyer, MIT, United States
Olga Troyanskaya, Princeton, United States
Jeff Leek, Johns Hopkins, United States
Ben Macarthur, Mount Sinai School of Medicine, United States
Alexander Lachmann, Mount Sinai School of Medicine, United States
Roye Rozov, Mount Sinai School of Medicine, United States
Avi Ma'ayan, Mount Sinai School of Medicine, United States
Ihor Lemischka, Mount Sinai School of Medicine, United States

Presentation Overview:
Molecular regulation of embryonic stem cell (ESC) fate involves a coordinated interaction between epigenetic, transcriptional and translational mechanisms. It is unclear how these different molecular regulatory mechanisms interact to regulate changes in stem cell fate. Here we present a dynamic systems-level study of cell fate change in murine ESCs following a well-defined perturbation. Global changes in histone acetylation, chromatin-bound RNA polymerase II, messenger RNA, and nuclear protein levels were measured over 5 days after downregulation of Nanog, a key pluripotency regulator. Our data demonstrate how a single perturbation leads to progressive changes in several molecular regulatory layers, and provide a dynamic view of information flow in the epigenome, transcriptome and proteome. The temporal order of gene expression alterations shows the order of the regulatory network reconfiguration and offers further insight into the gene regulatory network. Our studies underscore the complexity of the multilayered regulatory mechanisms that determine stem cell fate.

TOP

Comparative Network Analysis of Complex Diseases
Room: Ballroom A/B
Presenting author: Rune Linding, The Institute of Cancer Research, United Kingdom


Presentation Overview:
Insights into the evolution of protein phosphorylation were revealed by combining the results from two computational analysesÑa sequence-alignment approach and a kinase-substrate network alignment approach. The two approaches yielded different, but somewhat overlapping, sets of conserved phosphoproteins among humans and the model organisms. The first provided a set of genes encoding phosphoproteins that had positionally conserved phosphorylation sites, whereas the second included many functionally conserved phosphoproteins that lacked this positional conservation. Enrichment analysis of the genes identified through the kinase-substrate network approach suggested that genes encoding phosphorylated signaling hubs were enriched in disease-associated genes. Our analysis also suggests that conserved regulatory networks may be involved in different diseases. These findings may produce new targets for therapeutic intervention or permit researchers to predict the best combinations of therapeutics for intervening in diseases associated with aberrant signaling networks.

TOP

A systems approach shows competition and saturation drives microRNA and siRNA target gene regulation - its more than target site efficacy
Room: Ballroom C
Presenting author: Debora Marks, Harvard Medical School, United States

Additional authors:
aaron arvey, MSKCC, United States
Aly Khan, MSKCC, United States
Erik Larsson, MSKCC, United States
Christina Leslie, MSKCC, United States
Martin Miller, MSKCC, United States

Presentation Overview:
Saturation of protein machinery in RNAi pathways may affect both siRNA and microRNA targeting. By analyzing hundreds of experiment, we show that genes which are targets of endogenous micoRNAs are unexpectedly and unintentionally upregulated after microRNA/siRNA perturbations. We go on to discover over 20 novel significant motifs in 3'UTRs which work cooperatively with microRNAs to alter gene expression. Competition between different mRNAs for the microRNAs or siRNAs should, in theory affect the targeting quantitatively, according to a basic kinetic model. Taking hundreds of experiments we show that target abundance is a strong determinant in microRNA regulation.Our results show that mRNA target abundance and the competition for miRNAs and siRNAs has global consequences. This provides strong support for a re-assessment of what determines micro/siRNA targeting. Specifically, our results will shed insight into several critical problems to the community, including microRNA target prediction, siRNA screen design and small RNA therapeutics.

TOP

Systematic prediction of human membrane receptor interactions
Room: Ballroom A/B
Presenting author: Yanjun Qi, NEC Labs America, United States

Additional authors:
Harpreet Dhiman, University of Pittsburgh School of Medicine, United States
Neil Bhola, Department of Otolaryngology, United States
Ivan Budyak, Institute for Structural Biology, Germany
Siddhartha Kar, University of Pittsburgh School of Medicine, United States
David Man, University of Pittsburgh School of Medicine, United States
Arpana Dutta, University of Pittsburgh School of Medicine, United States
Kalyan Tirupula, University of Pittsburgh School of Medicine, United States
Brian Carr, University of Pittsburgh School of Medicine, United States
Jennifer Grandis, University of Pittsburgh School of Medicine, United States
Ziv Bar-Joseph, Carnegie Mellon University, United States
Judith Klein-Seetharaman, University of Pittsburgh School of Medicine, United States

Presentation Overview:
The paper presents a combined computational/experimental approach allowing us to determine, for the first time, the global network of human membrane receptor interactions. In addition to systematic analysis, the interactome generates very specific and experimentally verifiable predictions on individual pairwise interactions. We experimentally tested several novel predicted interactions for EGFR.

TOP

Determining the distribution of probes between different subcellular locations through automated unmixing of subcellular patterns
Room: Ballroom C
Presenting author: Robert Murphy, Carnegie Mellon University, United States

Additional authors:
Tao Peng, Carnegie Mellon University, United States
Ghislain Bonamy, Genomics Institute of the Novartis Research Foundation, United States
Estelle Glory-Afshar, Carnegie Mellon University, United States
Daniel Rines, Genomics Institute of the Novartis Research Foundation, United States
Sumit Chanda, Burnham Institute, United States

Presentation Overview:
Many proteins and macromolecules exhibit complex subcellular distributions, including localizing in more than one organelle and varying in location depending on cell physiology. Estimating the fraction of fluorescence in each organelle is essential to understanding protein dynamics and function. This paper describes the first validation of a machine learning approach, pattern unmixing, for estimating the amount of fluorescent signal in different subcellular compartments using only a single fluorescent probe. It does not require hand tuned image analysis algorithms and requires only the acquisition of separate training images of markers for each compartment. The method was validated using a collection of images of cells labeled with mixtures of different fluorescent probes that was specifically generated for this study and made publicly available. It will be useful for image-based proteome-scale localization tasks, such as drug effect monitoring and protein dynamics screening and interpretation.

TOP

Automating Biology Using Robot Scientists
Room: Ballroom A/B
Presenting author: Ross King, University of Wales - Aberystwyth, United Kingdom

Additional authors:
Jem Rowland, Aberystwyth University, United Kingdom
Wayne Aubrey, Aberystwyth University, United Kingdom
Maria Liakata, Aberystwyth University, United Kingdom
Larisa Soldatova, Aberystwyth University, United Kingdom
Whelan Kenneth, Aberystwyth University, United Kingdom
Clare Amanda, Aberystwyth University, United Kingdom
Sparkes Andrew, Aberystwyth University, United Kingdom
Michael Young, Aberystwyth University, United Kingdom
Magdalena Markham, Aberystwyth University, United Kingdom
Pınar Pir, Cambridge University, United Kingdom
Steve Oliver, Cambridge University, United Kingdom

Presentation Overview:
The basis of science is the hypothetico-deductive method and the recording of experiments in sufficient detail to enable reproducibility. We report the development of the Robot Scientist 'Adam' which advances the automation of both. Adam has autonomously generated functional genomics hypotheses about the yeast Saccharomyces cerevisiae and experimentally tested these hypotheses by using laboratory automation. We have confirmed Adam's conclusions through manual experiments. To describe Adam's research, we have developed an ontology and logical language. The resulting formalization involves over 10,000 different research units in a nested treelike structure, 10 levels deep, that relates the 6.6 million biomass measurements to their logical description. This formalization describes how a machine contributed to scientific knowledge.

TOP

Computational Models of the Notch Network Elucidate Mechanisms of Context-dependent Signaling
Room: Ballroom C
Presenting author: Smita Agrawal, University of Minnesota, United States

Additional authors:
Colin Archer, PhD Student, Australia
David Schaffer, Professor, United States

Presentation Overview:
The Notch signaling pathway has been implicated in numerous cell fate decisions both during development and adulthood. Extensive studies of this pathway indicate that the same core circuit functions in different cellular contexts to elicit varied behaviors and responses ranging from a cellular oscillator to a cell fate switch. Malfunctioning of this critical signaling pathway is also implicated in various cancers. To better understand the underlying mechanisms that allow the network to function distinctly in different contexts, we have developed deterministic and stochastic models to simulate the behavior of the Notch network. Our results indicate that behavior of the system can readily be tuned based on some key parameters to reflect its multiple roles in both the developing and adult organisms. Furthermore, the results provide insights into alterations in the signaling system that lead to malfunction and hence disease, which could be used to identify potential drug targets for therapy.

TOP

FragBag: representing protein structures as 'bags-of-fragments' allows efficient exploration of protein structure space.
Room: Ballroom A/B
Presenting author: Rachel Kolodny, University of Haifa, Israel

Additional authors:
Inbal Budowski-Tal, Graduate Student, Israel
Yuval Nov, Assistant Proffessor, Israel

Presentation Overview:
We present FragBag, a novel way to represent a protein structure as a 'bag-of-fragments', based on its short contiguous backbone segments. Using FragBag, we can quickly and accurately find candidate sets of structural neighbors of a query protein in a large database of structures. We validate the accuracy of our method with respect to a stringent gold standard, and show that it performs on a par with the computationally expensive, yet highly trusted, STRUCTAL and CE. Our representation has additional benefits: it can be used to construct an inverted index for implementing a fast structural search engine of the PDB, and one can specify a structure by its (uncombined) substructures; this is valuable for structure prediction, when there are reliable predictions only of parts of the protein. Finally, we show how to use FragBag to visualize protein structure space, and the insights that we gain from this visualization.

TOP

The Genetic Landscape of a Cell
Room: Ballroom C
Presenting author: Anastasia Baryshnikova, University of Toronto, Canada

Additional authors:
Michael Costanzo, University of Toronto, Canada
Jeremy Bellay, University of Minnesota, United States
Yungil Kim, University of Minnesota, United States
Eric D. Spear, Massachusetts Institute of Technology, United States
Carolyn S. Sevier, Massachusetts Institute of Technology, United States
Huiming Ding, University of Toronto, Canada
Judice L. Y. Koh, University of Toronto, Canada
Kiana Toufighi, University of Toronto, Canada
Sara Mostafavi, University of Toronto, Canada
Jeany Prinz, University of Toronto, Canada
Robert P. St. Onge, Stanford University, United States
Benjamin VanderSluis, University of Minnesota, United States
Taras Makhnevych, University of Toronto, Canada
Franco J. Vizeacoumar, University of Toronto, Canada
Solmaz Alizadeh, University of Toronto, Canada
Sondra Bahr, University of Toronto, Canada
Renee L. Brost, University of Toronto, Canada
Yiqun Chen, University of Toronto, Canada
Murat Cokol, Harvard Medical School, United States
Raamesh Deshpande, University of Minnesota, United States
Zhijian Li, University of Toronto, Canada
Zhen-Yuan Lin, Mount Sinai Hospital, Canada
Wendy Liang, University of Toronto, Canada
Michaela Marback, University of Toronto, Canada
Jadine Paw, University of Toronto, Canada
Bryan-Joseph San Luis, University of Toronto, Canada
Ermira Shuteriqi, University of Toronto, Canada
Amy Hin Yan Tong, University of Toronto, Canada
Nydia van Dyk, University of Toronto, Canada
Iain M. Wallace, University of Toronto, Canada
Joseph A. Whitney, University of Toronto, Canada
Matthew T. Weirauch, University of California, Santa Cruz, United States
Guoqing Zhong, University of Toronto, Canada
Hongwei Zhu, University of Toronto, Canada
Walid A. Houry, University of Toronto, Canada
Michael Brudno, University of Toronto, Canada
Sasan Ragibizadeh, S&P Robotics, Inc., Canada
Balazs Papp, Biological Research Center, hu
Csaba Pal, Biological Research Center, hu
Frederick P. Roth, Harvard Medical School, United States
Guri Giaever, University of Toronto, Canada
Corey Nislow, University of Toronto, Canada
Olga G. Troyanskaya, Princeton University, United States
Howard Bussey, McGill University, Canada
Gary D. Bader, University of Toronto, Canada
Anne-Claude Gingras, Mount Sinai Hospital, Canada
Quaid D. Morris, University of Toronto, Canada
Philip M. Kim, University of Toronto, Canada
Chris A. Kaiser, Massachusetts Institute of Technology, United States
Chad L. Myers, University of Minnesota, United States
Brenda J. Andrews, University of Toronto, Canada
Charles Boone, University of Toronto, Canada

Presentation Overview:
Understanding how genes of an organism interact with one another to produce complex phenotypes is a primary challenge in deciphering the functional organization of living cells and the genetic basis of disease. We examined 5.4 million gene-gene pairs in yeast Saccharomyces cerevisiae and generated the largest genetic interaction map to date, containing genetic interaction profiles for ~75% of all yeast genes. Our genetic interaction network reveals a functional map of the cell in which genes of similar biological processes cluster together in coherent subsets. Highly correlated genetic profiles delineate specific pathways and enable the prediction of novel gene function. We also demonstrate that unbiased genome-wide mapping of the genetic interactions provides a key for interpreting chemical-genetic interactions and identifying drug targets. Finally, we show that genetic interactions identify functional cross-connections between all bioprocesses and correlate with several gene attributes, which may be informative about genetic network hubs in other organisms.

TOP

A probabilistic approach to the design of interfaces in proteins with multiple partners: Tradeoff between stability and promiscuity
Room: Ballroom A/B
Presenting author: Menachem Fromer, The Hebrew University of Jerusalem, Israel

Additional authors:
Chen Yanover, Fred Hutchinson Cancer Research Center, United States
Michal Linial, The Hebrew University of Jerusalem, Israel
Julia Shifman, The Hebrew University of Jerusalem, Israel

Presentation Overview:
Traditionally, computational protein design efforts have been directed at calculating a single sequence predicted to fold to a particular target structure. Recently, however, a number of conceptual generalizations have been pursued, ranging from the use of backbone flexibility, off-rotamer side chain flexibility, negative design, multi-body potentials, conformational free energy, and prediction of sequence profiles. Here, our goal was to understand how protein sequences are constructed to be compatible with binding multiple partners with high affinity. Such information would shed light on both naturally occurring evolution and directed evolution of protein sequences with multiple functions. We find that simultaneously designing for additional native partners yields sequences that better match wild-type sequence profiles, thus emphasizing the importance of such strategies in nature. Interestingly, we find that only some degree of compromise is typically needed in order to permit interactions that are seemingly antagonistic.

TOP

A 3D Digtial Atlas of C. elegans and Its Application to Single-Cell Gene Expression and Cell Fate Analyses
Room: Ballroom C
Presenting author: Fuhui Long, Howard Hughes Medical Institute, United States

Additional authors:
Hanchuan Peng, HHMI, United States
Xiao Liu, Stanford Univ, United States
Stuart Kim, Stanford Univ, United States
Gene Myers, HHMI, United States

Presentation Overview:
We built a digital nuclear atlas of the first larval stage (L1) of Caenorhabditis elegans at single-cell resolution from confocal image stacks. For the first time, the atlas quantifies the stereotypy of nuclear locations and provides other statistics on the spatial patterns of the 357 nuclei that could be faithfully segmented and annotated out of the 558 present at this developmental stage. We then developed an automated approach to assign cell names to each nucleus in a three- dimensional image of an l1 worm. This computational method will allow high-throughput single-cell analyses of the post-embryonic worm, such as gene expression analysis, or ablation or stimulation of cells under computer control in a high-throughput functional screen. We have applied this method to produce the single-cell resolution expression of 93 genes. Analyses indicated previously unknown new cell sub-types and interesting relationship between gene expression and cell fate.

TOP

Protein Folding Requires Crowd Control in a Simulated Cell
Room: Ballroom A/B
Presenting author: Benjamin Jefferys, Imperial College London, United Kingdom


Presentation Overview:
Macromolecular crowding has a profound effect upon biochemical processes in the cell. We have computationally studied the effect of crowding upon protein folding for 12 small domains in a simulated cell using a coarse-grained protein model, which is based upon Langevin dynamics, designed to unify the often disjoint goals of protein folding simulation and structure prediction. We found that when crowding approaches 40% excluded volume, the maximum level found in the cell, proteins fold to fewer native-like states. Notably, when crowding is increased beyond this level, there is a sudden failure of protein folding: proteins fix upon a structure more quickly and become trapped in extended conformations. These results suggest that the ability of small protein domains to fold without the help of chaperones may be an important factor in limiting the degree of macromolecular crowding in the cell.

TOP

Protein interactions and ligand binding: from protein subfamilies to functional specificity
Room: Ballroom A/B
Presenting author: Antonio Rausell, Spanish National Cancer Research Centre (CNIO), Spain

Additional authors:
David Juan, Spanish National Cancer Research Centre (CNIO), Spain
Florencio Pazos, National Centre for Biotechnology (CNB-CSIC), Spain
Alfonso Valencia, Spanish National Cancer Research Centre (CNIO), Spain

Presentation Overview:
The divergence accumulated during the evolution of protein families translates into their internal organization as subfamilies, and it is reflected in the characteristic patterns of differentially conserved residues. These specifically conserved positions in protein subfamilies are known as 'specificity determining positions' (SDPs). Previous studies have limited their analysis to the study of the relationship between these positions and ligand-binding specificity. We have systematically extended this observation to include the role of differential protein interactions in the segregation of protein subfamilies and explored in detail the structural distribution of SDPs at protein interfaces. Our results show the extensive influence of protein interactions in the evolution of protein families and the widespread association of SDPs with interfaces. The combined analysis of SDPs in interfaces and ligand-binding sites provides a more complete picture of the organization of protein families, constituting the necessary framework for a large-scale analysis of the evolution of protein function.

TOP

Three-Dimensional Structural View of the Central Metabolic Network of Thermotoga maritima
Room: Ballroom A/B
Presenting author: Ying Zhang, Sanford-Burnham Medical Research Institute, United States

Additional authors:
Ines Thiele, University of Iceland, Iceland
Dana Weekes, Sanford-Burnham Medical Research Institute, United States
Zhanwen Li, Sanford-Burnham Medical Research Institute, United States
Lukasz Jaroszewski, Sanford-Burnham Medical Research Institute, United States
Krzysztof Ginalski, Warsaw University, pl
Ashley Deacon, Joint Center for Structural Genomics (JCSG), United States
John Wooley, University of California at San Diego, United States
Scott Lesley, Genomics Institute of the Novartis Research Foundation, United States
Ian Wilson, The Scripps Research Institute, United States
Bernhard Palsson, University of California at San Diego, United States
Andrei Osterman, Sanford-Burnham Medical Research Institute, United States
Adam Godzik, Sanford-Burnham Medical Research Institute, United States

Presentation Overview:
Sequencing of a genome provides a foundation for the computational and experimental study of complete biological networks. Knowledge of three-dimensional structures of proteins that compose such networks extends our understanding of the biological processes to the atomic level. Here we developed a metabolic reconstruction for a hyperthermophilic bacterium, Thermotoga maritima, and combined it with 120 experimental and 358 computational structural models to achieve a 100% structural coverage of the proteins involved. Structural modeling of all proteins in a metabolic network opens new possibilities for simulations leading to deeper understanding of the function, mechanism, and evolution of metabolic pathways. Using information from the combined study of structural genomics and systems biology, we develop tools to validate the metabolic reconstruction and extend the scope of this in silico model by predicting the function of uncharacterized proteins. These tools will enable the reconstruction of a living model of a complete T. maritima cell.

TOP

An Integrative Proteomics Approach to Identify Functional Sub-networks In Cancer
Room: Ballroom A/B
Presenting author: Rod Nibbe, Case Western Reserve University, United States

Additional authors:
Mehmet Koyuturk, Case Western Reserve University, United States
Mark Chance, Case Western Reserve University, United States

Presentation Overview:
Sub-networks within PPI networks that demonstrate synergistic differential expression with respect to tumorigenic phenotypes were recently shown to be more accurate classifiers of metastasis when compared to single targets identified by traditional approaches. We show that proteomic targets significant for human colorectal cancer (CRC) may be used to "seed" a novel, global network search to discover sub-networks functionally associated with these targets. We evaluate synergistic changes in the activity of these sub-networks (tumor vs. control) based on genome-wide screens of mRNA expression in CRC. We show that sub-networks with significant crosstalk to proteomic targets are synergistically differentially expressed at the transcriptional level. Proteomic targets are statistically as powerful in discovering such sub-networks as a seed of candidate driver genes of CRC. Further, we validate by wet-bench experiment the functional role of select sub-network targets in CRC and show how they are useful as features to train a highly accurate classifier.

TOP

Fast and accurate large-scale co-estimation of alignments and trees
Room: Ballroom A/B
Presenting author: Tandy Warnow, University of Texas at Austin, United States

Additional authors:
Kevin Liu, UT-Austin, United States
Sindhu Raghavan, UT-Austin, United States
Serita Nelesen, UT-Austin, United States
C. Randal Linder, UT-Austin, United States

Presentation Overview:
I will present SATe (Simultaneous Alignment and Tree Estimation), a method that uses a novel re-alignment technique to iteratively estimate alignments and trees. Our study shows dramatic improvements in accuracy over all existing methods, with 24 hours sufficing for 1000 taxon datasets. I will also present unpublished results for a more accurate method, SATe-2.

TOP

Genome-Wide Association Data Reveal a Global Map of Genetic Interactions among Protein Complexes
Room: Ballroom C
Presenting author: Rohith Srivas, University of California, San Diego, United States

Additional authors:
Gregory Hannum, University of California, San Diego, United States
Aude Guénolé, Leiden University Medical Center, Netherlands
Haico van Attikum, Leiden University Medical Center, Netherlands
Nevan Krogan, University of California, San Francisco, United States
Richard Karp, University of California, Berkeley, United States
Trey Ideker, University of California, San Diego, United States

Presentation Overview:
Despite the immense potential of gene association studies, they have been challenging to analyze because most traits are complex, involving the combined effect of mutations at many different genes. Due to lack of statistical power, only the strongest single markers are typically identified. Here, we present an integrative approach that greatly increases power through marker clustering and projection of marker interactions within and across protein complexes. Applied to a recent gene association study in yeast, this approach identifies 2,023 genetic interactions which map to 208 functional interactions among protein complexes. We show that such interactions are analogous to interactions derived through reverse genetic screens and that they provide coverage in areas not yet tested by reverse genetic analysis. As proof of principle, we use synthetic genetic screens to confirm numerous novel genetic interactions for the INO80 chromatin remodeling complex.

TOP

Selection for hotspot mutations in core genes of Escherichia coli
Room: Ballroom A/B
Presenting author: Sujay Chattopadhyay, University of Washington, United States

Additional authors:
Scott Weissman, University of Washington, United States
Vladimir Minin, University of Washington, United States
Thomas Russo, SUNY Buffalo, United States
Daniel Dykhuizen, SUNY Stony Brook, United States
Evgeni Sokurenko, University of Washington, United States

Presentation Overview:
Core genes comprising the ubiquitous backbone of bacterial genomes are not subject to frequent horizontal transfer and generally are not thought to contribute to the adaptive evolution of bacterial pathogens. We determined, however, that at least a third of the core genes in Escherichia coli genomes are targeted by repeated replacement substitutions in the same amino acid positions Ð hotspot mutations. Occurrence of hotspot mutations is driven by positive selection, as their rate is significantly higher than random expectations, and neither recombination nor increased mutability can explain the observed patterns. Pathogenic E. coli strains, as opposed to non-pathogenic ones, accumulate hotspot mutations at significantly higher rate, suggesting the pathoadaptive nature of such mutations. Vast majority of hotspot mutations are of recent evolutionary origin, implying short-term positive selection, where adaptive mutations emerge repeatedly but are not sustained in natural circulation for long (a dynamics consistent with source-sink model of virulence evolution).

TOP

Network models for understanding what 'omic data really mean.
Room: Ballroom C
Presenting author: Ernest Fraenkel, Massachusetts Institute for Technology, United States


Presentation Overview:
Typically, most of the hits identified by high-throughput assays fall outside of the expected pathways. These unexpected components of the cellular response are often the most interesting, because they can provide new insights into biological processes. However, they are also the most difficult to interpret.
We provide a method for identifying unexpected components of cellular responses by linking diverse data including phosphoproteomic, genetic and transcriptional data through previously reported protein-protein and protein-DNA interactions. Our approach, which is based on the Prize collecting Steiner Tree problem identifies compact networks containing functionally coherent pathways, revealing components of the cellular response that are not readily apparent in the original data.

TOP

Neofunctionalization in interaction network evolution
Room: Ballroom A/B
Presenting author: Todd Gibson, University of Colorado Denver, United States

Additional authors:
Debra Goldberg, University of Colorado, United States

Presentation Overview:
Analyses of gene duplication and link dynamics underlie studies of protein interaction network evolution. Here we revisit assays and methods used is this research. We show the importance of analyzing duplication events which have shaped the extant empirical interactions, not as isolated events, but rather in concert with concurrent and subsequent duplication events. We also revisit the Y2H and AP-MS assays which generate the bulk of empirical data upon which protein interaction network research is based. We note that both of these assays are biased against reporting self-interacting proteins, leaving a dearth of homomeric interactions in empirical data sets. Finally, we show that the self-interactions largely overlooked by these assays are integral to the high clustering observed in the empirical data...We examine the impact these phenomena have on the network literature, and find methodological oversights in studies which have found that neofunctionalization (the gain of new function by gene duplicates) is prevalent in protein interaction network evolution.

TOP

Cell-specific information processing in segregating populations of Eph receptor ephrin-expressing cells.
Room: Ballroom C
Presenting author: Claus Jørgensen, The Institute of Cancer Research, United Kingdom


Presentation Overview:
Here we presented the first quantitative network model of cell specific Eph receptor-ephrin signal processing following cell to cell contact. Cell specific phospho-tyrosine signaling was measured between Eph- and ephrin-expressing cells using a novel quantitative mass spectrometric approach, which subsequently was integrated with phenotypic data of receptor-ligand function, obtained by siRNA screening. Through data-driven computational modeling, cell specific networks were constructed to describe quantitative signaling trajectories from kinases to effecter modules. Comparative analysis between receptor and ligand expressing cells revealed cell specific differences in the network utilization, in part explained by differences in kinase activity. ..Network analysis of signaling trajectories between wild-type and signaling impaired Eph-ephrin mutants provided insight into underlying differences in the networks controlling cell specific behavior, and suggested non-cell autonomous effects of both the Eph receptor and the ephrin. Finally, we show that signaling trajectories were utilized significantly different dependent on the context.

TOP

A universal relationship between gene compactness and expression level in multicellular eukaryotes
Room: Ballroom A/B
Presenting author: Liran Carmel, The Hebrew University of Jerusalem, Israel

Additional authors:
Eugene Koonin, National Institutes of Health, United States

Presentation Overview:
Are highly expressed genes more compact? Evidence suggests that the answer is positive for animals, but negative for plants. Examining this intriguing question in four organisms (A. thaliana, human, worm, and fly), we identify a universal nonmonotonic relationship: highly expressed genes tend to be shorter, indeed, but genes with very low expression level are almost as short. At intermediate expression levels, genes are the longest. The biology that underlies this relationship is not clear. We discuss several possible explanations, and conclude that highly expressed genes are likely to be shorter because of selection pressure (for instance, selection for minimization of energy and time expenditure during transcription and splicing and for increased fidelity of transcription, splicing, and/or translation). Regardless of the exact nature of the forces that shape the gene architecture, we present evidence that, at least, in animals, coding and noncoding parts of genes show similar architectonic trends.

TOP

Pathway and network based approaches to prioritize reliable hits from high throughput RNAi screening experiments
Room: Ballroom C
Presenting author: Zhidong Tu, Merck, United States

Additional authors:
Li Wang, University of Southern California, United States
Fengzhu Sun, University of Southern California, United States
Carmen Argmann, Merck, United States
Kenny Wong, Merck, United States
Lyndon Mitnaul, Merck, United States
Stephen Edwards, Merck, United States
Iliana Sach, Merck, United States
Jun Zhu, Merck, United States
Eric Schadt, Merck, United States

Presentation Overview:
Not Available

TOP

Scaling to PubMedCentral: Bringing Ontology-based Biomedical Language Processing into the Full Text Era
Room: Ballroom A/B
Presenting author: Lawrence Hunter, University of Colorado School of Medicine, United States


Presentation Overview:
As full texts of biomedical journal publications become increasingly available, new tools and engineering methodologies are necessary to take full advantage of this extraordinarily valuable resource. Recent work clarifies many of the challenges, as well as demonstrating the value of building on existing ontologies and middleware frameworks.

TOP

Applying histone modification information to genome-wide prediction of transcription factor binding sites
Room: Ballroom C
Presenting author: Kyoungjae Won, University Of California, San Diego, United States

Additional authors:
Bing Ren, UCSD, United States
Wei Wang, UCSD, United States

Presentation Overview:
We present an integrated method called Chromia for the genome-wide identification of functional target loci of transcription factors. Designed to capture the characteristic patterns of a transcription factor binding motif occurrences and the histone profiles associated with regulatory elements such as promoters and enhancers, Chromia significantly outperforms other methods in the identification of 13 transcription factor binding sites in mouse embryonic stem cells, evaluated by both binding (ChIP-seq) and functional (RNAi knockdown) experiments. Capturing characteristics of multi-dimensional epigenetic signatures within the states of hidden Markov models, Chromia, identified functional binding sites associated with histone signatures.

TOP

Modeling the temporal interplay of molecular signalling and gene expression by using dynamic nested effects models
Room: Ballroom A/B
Presenting author: Benedict Anchang, University of Regensburg, Germany

Additional authors:
Rainer Spang, University of Regensburg, Germany
Peter Oefner, University of Regensburg, Germany
Mohammad Sadeh, University of Regensburg, Germany
Juby Jacob, University of Regensburg, Germany
Achim Tresch, Ludwig-Maximilians-Universitaet Muenchen, Germany

Presentation Overview:
Cellular decision making in differentiation, proliferation or cell death is mediated by molecular signaling processes, which control the regulation and expression of genes. Changes in gene expression can activate further signaling processes, leading to secondary effects, which themselves give rise to tertiary effects and so on. The result is an intricate interplay of cell signaling and gene regulation. While protein modification in the cytoplasm can propagate signals in seconds, transcription and translation processes last hours, and secondary effects often become visible only after days. We developed a statistical method called Dynamic Nested Effects Model (D-NEM) for analyzing the temporal interplay of cell signaling and gene expression. In an application to decision making in mural embryonic stem cell development, we could show that a feed-forward loop dominated gene regulation network ensures that cell differentiation is a quasi unidirectional process in vivo. A reversal of the differentiation process would cause a latent cancer risk.

TOP

Histone modification levels are predictive for gene expression
Room: Ballroom C
Presenting author: Martin Vingron, Max-Planck-Institut fur Molekulare Genetik, Germany

Additional authors:
Rosa Karlic, Max-Planck-Institut fur Molekulare Genetik, Germany
Ho-Ryun Chung, Max-Planck-Institut fur Molekulare Genetik, Germany
Julia Lasserre, Max-Planck-Institut fur Molekulare Genetik, Germany
Kristian Vlahovicek, Faculty of Science, Zagreb University, hr

Presentation Overview:
Histones are frequently decorated with covalent modifications, which are involved in various chromatin-dependent processes. To elucidate the relationship between histone modifications and transcription, we derived quantitative models to predict expression levels of genes from histone modification levels. We found that histone modification levels and gene expression are highly correlated and showed that only a small number of histone modifications are necessary to accurately predict gene expression. We show that different sets of histone modifications are necessary to predict gene expression driven by high CpG content promoters (HCPs) or low CpG content promoters (LCPs). Quantitative models using H3K4me3 and H3K79me1 are the most predictive for expression of LCPs, whereas HCPs require H3K27ac and H4K20me1. We show that the connections between histone modifications and gene expression seem to be general, as we were able to predict gene expression levels of one cell type using a model trained on another one.

TOP

Illuminating Complete Functional Networks: Automation, Computation and the Single Cell
Room: Ballroom A/B
Presenting author: Michael Fero, Stanford University, United States

Additional authors:
Beat Christen, Stanford, United States

Presentation Overview:
Advances in high resolution microscopy and quantitative analysis of subcellular protein location and abundance have greatly enriched number and quality of phenotypic measurements accessible to the biologist. Now, with the development of high speed data acquisition and automated analysis, single subcellular level data can be gathered in the context of saturated genetic screens, yielding rich data sets that can be used to associate genotypic disruptions or perturbations with clear subcellular phenotypes. Modern screening techniques allow us to probe gene function on a genome scale, from engineered transposons in prokaryotic systems, to systematic knock-outs in Saccharomyces cerevisiae, to RNA interference (RNAi) knock-down libraries in a variety of organisms By combining rich subcellular phenotypic data with known gene perturbations on a genome scale, we can now identify key functional networks including signaling, transcription and protein localization factors in a single genetic screen.

TOP

Modeling information flow from metabolomics to transcriptomics
Room: Ballroom C
Presenting author: Xinghua Lu, Medical University of South Carolina, United States

Additional authors:
L. Ashley Cowart, Medical University SC, United States
Adam Richards, Medical University SC, United States
David Montefusco, Medical University SC, United States
Yusuf Hannun, Medical University SC, United States
Matthews Shotwell, Medical University SC, United States

Presentation Overview:
We present a systems-biology approach recently developed to model the information flow from cellular stress to metabolic changes then further to transcriptomic changes. We studied a yeast system in which cellular stresses cause changed metabolism of a family of bioactive lipids referred to as sphingolipids, which further mediate gene expression in response to cellular stress. We have developed a statistical framework which, through integrating multiple types of high throughput data, infers the activation states of transcription factors (TFs), reveals the information (connectivity) between lipidomic and transcriptomic data, and models activation of TFs by specific lipids. Our model led to a testable hypothesis revealing a signal transduction pathway involving phytospingosine-1-phosphate to HAP complex and then to genes involved in energy metabolism. The hypothesis was experimentally validated.

TOP