Frank Alber (Molecular and Computational Biology, University of Southern California, Los Angeles, CA 90089)

Title: Exploring the Nuclear Architecture: The Structural Characterization of the Nuclear Pore Complex and the Genome.

Abstract: To understand the workings of the living cell, we need a detailed description of the architectures of its macromolecular assemblies. The broad aim of our research is to develop and apply a system for determining how macromolecular complexes structurally assemble and accomplish complex cellular processes. Our approach will integrate structural information gathered at multiple levels of the biological hierarchy-from atoms to cells-into a common framework. We exemplify our approach by exploring the key elements of the nuclear architecture.
The first part of the talk will focus on how integration of low-resolution proteomic and biophysical data can be used to determine the subunit architecture of the nuclear pore complex (NPC). The NPC acts as a dynamic barrier to control access to and from the nucleus, and in yeast is a 50 MDa assembly of 456 proteins. The structure determination process involves collection of sufficient and diverse proteomic and biophysical data, translation of this data into spatial restraints, and an optimization that uses these restraints to generate an ensemble of structures consistent with the data. Analysis of the ensemble produces a detailed architectural map and interaction network of the assembly. Our resulting structure reveals the configuration of the proteins in the NPC, and provides insights into the evolution and architectural principles.
In a second line of research, we apply our integrated hierarchical system to characterize the higher order structure of the interphase genome by exploiting all available relevant information about its composition and structure. Our resulting structure reveals organizational principles that guide the genome organization.

Kiyoshi Asai (University of Tokyo & Computational Biology Research Center, AIST)

Title: Estimation in High-dimensional Binary Space for RNA Informatics

Abstract: Recent findings revealed that non-coding functional RNAs play important role in the regulation of the cell mechanisms. For the sequences of non-coding RNAs, popular software tools of sequence analyses produce appropriate results neither in homology searches nor in alignments. It is known that considering the secondary (2D) structures of RNAs as well as the sequences themselves is the key for the analyses. Consideration of 2D structures, however, can be too expensive for genome-scale sequence analyses. Strict solution for the alignment of two RNA sequences (length L) consistent to their 2D structures, for example, requires O(L6) in time. Recent progress in theories and algorithms, however, has changed the situation. We have been contributed in various problems for analyses of RNA sequences, such as predictions of 2D structures, 2D structural alignments and common 2D structure extractions. In this talk, various problems in biological sequence analyses are formulated as estimation problems in high-dimensional binary space, and it is shown that those problems are solved by maximizing the expected value for appropriately designed gain functions of the problems.

Keith C.C. Chan (Department of Computing, The Hong Kong Polytechnic University)

Title: Towards Automatic Molecular Design: Using Evolutionary and Graph Mining Algorithms

Abstract: Computer-Aided Molecular Design (CAMD) is concerned with the use of computational techniques to determine molecular structures with certain desirable properties. Traditionally, the problem is tackled with the use of heuristic search or mathematical programming techniques but these techniques do not handle large and nonlinear search space very well. To overcome these drawbacks, evolutionary algorithms (EAs) have been proposed to evolve molecular design by mimicking chemical reactions that cause the exchange of chemical bonds and components between molecules. For these EAs to perform their tasks, known molecular components, which can serve as building blocks for the molecules to be designed, and known chemical rules, which govern chemical combination between different components, have to be introduced before the evolutionary process can take place. To automate molecular design without such prior knowledge and constraints, we need a special EA that can evolve molecular graphs with minimal background knowledge. In this talk, we discuss the specific requirements for such an EA and explain how such an EA can make use of graph mining algorithms to generate initial populations of molecular graphs for evolution and to evaluate the goodness of the designs obtained at each cycle of the evolutionary process. Results based on preliminary work to develop such an EA for automatic drug design will be presented. We explain how the results seem to indicate that such an approach can be very promising.

Luonan Chen (Key Laboratory of Systems Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences)

Title: Network-based Ontology Analysis

Abstract: Gene ontology based functional analysis is a major bioinformatics initiative to bridge the genotype and phenotype by systematically extracting biological meanings from an integrated, standardized, and cross-species annotation platform. The traditional gene ontology analysis is mainly conducted in individual genes or a gene list. However, recent network analysis reveals that the same set of genes with different interactions may perform different functions, i.e., network rewiring is able to facilitate the function changes between conditions. Therefore, it is necessary to annotate the specific functions of networks by considering the fundamental roles of interactions. Here, we develop a novel functional enrichment analysis method by focusing on the network, i.e., Network-based gene Ontology Analysis (NOA). Specifically, NOA defines link ontology to assign functions to links and then generates two alternative reference sets to statistically rank the enriched functional terms for a given biological network. We compared NOA with traditional gene list based functional annotation tools in several biological networks, and found that 1) NOA can find more relevant and specific functions; 2) NOA can capture the changing functions by network rewiring which significantly outperforms the gene list methods. In addition, NOA is developed as a web server and freely accessible at http://www.aporc.org/noa/.

Kwang-Hyun Cho (Department of Bio and Brain Engineering Korea Advanced Institute of Science and Technology (KAIST), Daejeon 305-701, Republic of Korea)

Title: State Space Analysis of Cell Fate Decision Dynamics

Abstract: Cell fate decision mechanism involves many complex regulatory steps, but we can conceptualize it as a state transition process over a genetic state space. In this framework, the cell fate decision depends on the energy landscape of the underlying genetic state space and cancer cell development can be considered as being initiated by lowering an energy barrier in the landscape and being progressed by converging to an unexpected cancer attractor. These concepts are largely based on the well-known 'epigenetic landscape' proposed by Waddington in 1950s. In this presentation, I show how we can reincarnate such an old concept to understand the complex regulatory mechanism of cell fate decision and cancer cell development within the context of systems biology. In particular, state space analysis of p53 regulatory dynamics will be used as a vehicle for discussion.

Wen-Lian Hsu (Institute of Information Science, Academia Sinica, Taipei)

Title: Protein Structure and Function Prediction

Abstract: We shall illustrate several techniques attacking protein structure as well as function prediction problems. For structure prediction, we shall focus on secondary structure prediction, and membrane helical structure prediction. For function prediction, we shall consider the related subcellular localization prediction. Also, we shall demonstrate a general knowledge-based approach for both structure and function prediction.

Sunghoon Kim (WCU Department of Molecular Medicine and Biopharmaceutical Sciences Medicinal Bioconvergence Research Center, Seoul National University, Korea)

Title: Innovation of Drug Discovery through Bioconvergence Technology

Abstract: Decrease of R&D productivity in drug discovery is a global problem and throws a great challenge not only to pharmaceutical industry but also to basic bioscience. The “BIOCON” project was designed to provide various innovative tools to facilitate the current difficulties in drug discovery. The project consists of the creative technologies for drug design, target identification, drug screening and disease modeling. The collection of these techniques will be linked as integrated form that will be conducted in integrated form. Various cutting-edge technologies in information, nano, chemistry, physics and medicine will be converged to address challenging questions in basic biology as well as to facilitate drug research. This presentation would introduce the basic concept, vision and goal of the “BIOCON” project.

Lei Liu

Title: Protein Phosphorylation Plays an Essential Role in the Evolution of Vertebrates

Abstract Recent publications have revealed that the evolution of phosphosites are influenced by local protein structures and whether the phosphosites have characterized functions or not. With knowledge of the wide functional range of phosphorylation, we attempted to clarify whether the evolutionary conservation of phosphosites is different among distinct functional modules. We grouped the phosphosites in the human genome into modules according to the functional categories of KEGG, and we investigated their evolutionary conservation in several vertebrate genomes from mouse to zebrafish. We found that phosphosites in the basic functional modules (BFMs) such as metabolic and genetic processes, display a lower evolutionary conservation than those in the vertebrate-specific functional modules (VFMs) such as signaling processes and more complex organic systems. The phosphosites in the VFMs are also significantly more conserved than their flanking regions, but those in the BFMs are not. The above results hold for both serine/threonine and tyrosine residues, although the fraction of phosphorylated tyrosine is raised in the VFMs. Moreover, the difference in evolutionary conservation of the phosphosites cannot be explained by their difference in local protein structures, and there are also more phosphosites with known functions in the VFMs. Based on these results, we concluded that protein phosphorylation may play more dominant roles for the VFMs than the BFMs. As phosphorylation is a quite rapid biological reaction, the VFMs that quickly respond to outer stimuli and inner signals might more heavily depend on this regulatory mechanism. Our results imply that phosphorylation may have an essential role in the evolution of vertebrates.

Y.M. Dennis Lo (Li Ka Shing Institute of Health Sciences and Department of Chemical Pathology, The Chinese University of Hong Kong, Shatin, New Territories, Hong Kong SAR, China)

Title: Noninvasive fetal genomic analysis by maternal plasma DNA analysis

Abstract: In 1997, our group reported the presence of cell-free fetal DNA in maternal plasma. This finding has opened up new possibilities for non-invasive prenatal diagnosis. Early work had focused on the detection of paternally-inherited genetic markers which were absent in the mother's genome, e.g. Y chromosomal markers. The development of next-generation sequencing has significantly expanded the diagnostic spectrum of plasma DNA based molecular diagnostics. Thus, fetal trisomy 21 can be robustly detected from maternal plasma using massively parallel sequencing. It is expected that this approach would also be useful for other chromosomal abnormalities and monogenic disorders. Finally, plasma DNA sequencing is also a powerful method for elucidating the biological characteristics of plasma nucleic acids, and might shed valuable insights into the mechanisms of its origin.

Tohru Natsume (Biomedicinal Information Research Center (BIRC), National Institute of Advanced Industrial Science and Technology (AIST), Japan)

Title: Challenge to Ultra-high Sensitive Mass Spectrometry for Protein Network Analysis

Abstract: Functional proteomics aims to discover gene functions at the protein level. Mass spectrometry (MS)-based proteomic approaches are powerful tools for identification and quantification analysis of the protein components of complexes in cells 1,2. Therefore, a variety of sample preparation techniques for the MS approach has been developed and reported3. Although these affinity purification methods are indeed useful, the problem is that the purified samples for proteome-wide analysis tend to be uneven in quality because of human elements. In large-scale protein interaction analysis, even though using partially-automated system, researchers or technicians are involved in laborious repetitive work of sample preparation, in which they must handle tens of culture dishes at a time and perform sequentially cell harvest and extraction, affinity purification of protein complexes, and enzymatic protein digestion. During the preparation of a number of samples, there must be difference in conditions between the first and last treated samples. Denaturation of the component proteins of the complexes and proteolysis are progressive over time, and the denaturated proteins are thought to be the cause of non-specifically binding. Furthermore, the difference of the sample quality is also caused by the personal skills.
To solve the problem, we have developed an automated sample preparation system for mass spectrometry-based functional proteomics. Since this robotic affinity purification system enables the isolation of protein complexes at short times under equal mild conditions, we successfully identify minor proteins reproducibly with non-specific proteins minimized. Using the well-characterized Wnt signaling pathway proteins, ?-catenin and Axin1, we could identify interaction partners, including novel component proteins, in mammalian cells.

Wing-Kin Sung (National University of Singapore)

Title: Bioinformatics Applications on Pathology

Abstract: Pathogens can cause diseases. They are dangerous since they can spread and mutate rapidly. To avoid pandemic, we need good methods for pathogen diagnosis and good methods to track the evoluation of pathogens. In this presentation, we will discuss our current works in these two aspects. We will first discuss how do we detect pathogens unbiasedly using microarray. Then, we discuss how do we track the mutations of viruses and bacteria using microarray and paired-end short reads, respectively.

Xuegong Zhang (Tsinghua University)

Title: Estimating Gene Expression Values from RNA-seq Data

Abstract: Studying the expression of genes at mRNA level is a basic step for many biological studies. The fastly advancing next generation sequencing technology has opened a new page of measuring RNA aoundance with deep sequencing (RNA-seq). RNA-seq has revealed many new observations on the characterics of mRNA expression, which makes the estimation of gene expression values not a straightforward task. In this talk, I'll report our efforts and observations on estimating gene expression values at isoform level and gene level for alternatively spliced genes.