Gultekin Ozsoyoglu - US grants

Workstations will be provided for researchers at Case Western Reserve University for research in the Department of Computer Engineering and Science. This equipment is provided under the Instrumentation Grants for Research and Information Science and Engineering program. The research which the equipment is to be used will be in the area(s) of Nested Relational Database Management System, Parallelism in Logic Programs and Synthetic Query Response Construction in Realtime/Time- Constrained Databases

Present database systems process all the data related to a query before giving out responses. As a result, the size of the data to be processed becomes excessive for time-constrained (and, perhaps, realtime) database application environments. For example, for Scientific database applications, data gathering and analysis capabilities can be greatly enhanced if there are real- time/online querying capabilities that return approximate responses to check the quality of the generated scienctific data. Also, with real-time/online "on-the-fly" data analysis (querying) capabilities, the size of the stored data in scientific databases can be kept small and, thus, manageable. Similar problems occur in time-constrained manufacturing database applications. A new methodology is needed to cut down systematically the time to process the data involved in processing the query. To this end, one can use data samples and utilize statistical approximation to construct an approximate response to given query. In this research, a methodology will be developed for the construction of approximate query responses for databases under time (and/or accuracy) constraints, and with incomplete data.

One may define a "time-constrained database" as a database that has strict timing constraints in (a) responding to queries, (b) processing transactions, and in (c) database maintenance. A "time-constrained query" has the form of "get the information x in no more than t time units". This project deals with the problem of processing arbitrary time-constrained queries in centralized and single-user time-constrained databases. Various ways of revising a query into one which has the most "informative" response and can be evaluated within the given time constraint with a controlled risk of overspending are investigated. Techniques for fragment selection and risk analysis are introduced. Semantic query processing techniques to control the query processing times are investigated. The expected outcome of this research is (i) a methodology to process queries with time constraints, (ii) a prototype time- constrained database management system and a performance evaluation using the prototype, and (iii) heuristics about the use of various options under different data characteristics. Potential applications of the research results are in those areas where time-critical database processing is important, e.g., scientific applications and manufacturing environments.

9224660 Ozsoyoglu SDB: Scientific Database Research for Polymers and Materials Engineering Needs This is the first year funding of a four-year continuing award. This research project develops a system for managing data generated by fracture study experiments in Polymers and Materials Engineering. The experiment data produced by materials engineers is large and has several data types that include summary data, image data, spatial data, and temporal data obtained by applying differing amounts of multiaxial stresses to materials at different times. The types of queries on the experiment data include statistical queries, matching queries, mapping queries, and temporal queries. The project involves integrating technologies from spatial and object-oriented databases, computational geometry, and image processing. New data structures, data models, query languages and query processing techniques that enable efficient storage and retrieval of data are designed. And a prototype scientific database management system is developed and tested. Such an integrated data management environment will benefit the researchers in polymers and materials engineering disciplines. Furthermore, the essential principles abstracted in this project can be used for data management in other disciplines. ***

CDA-9529503 Gultekin Ozsoyoglu Z. Meral Ozsoyoglu Case Western Reserve University Presentation Languages in Multimedia Databases Multimedia presentation refers to the concurrent presentation of multimedia data using monitors for text, video or still images, and speakers for audio. Each presentation contains multimedia segments retrieved from a database, and is constructed and played out by a presentation manager software. This research has three components. (a) A prototype system is developed to automate the assembly of presentations from multimedia databases for beginning computer users. The system is then revised and evaluated in different application domains. One example is a Course-Assistant software in education. (b) A presentation construction language for intermediate computer users is implemented and evaluated in different application domains. The language provides control on the number and contents of concurrent segments in a presentation, and has the ability to specify playout-time changes in a presentation, and communication and synchronization among multiple playout managers. (c) For advanced users, a presentation manipulation language is developed and evaluated. The development is carried out on multimedia personal computers with MPEG capture and playout capabilities and on workstations, acquired by the instrumentation grant. The development platforms are Windows NT and UNIX.

9912229
Ozsoyoglu

Description: This project supports a cooperative project by the Professor Gultekin Ozsoyoglu, Department of Electrical Engineering and Computer Science research at Case Western Reserve University (CWRU), Cleveland, Ohio and Dr. Ozgur Ulusoy, Department of Computer Engineering and Information Science, Bilkent University, Ankara, Turkey. They plan to investigate query languages for XML, the Extensible Markup Language which is proposed as a standard web Markup language by the World Wide Web Consortium. The research has two goals: to study ways of storing XML documents into databases, and querying such databases; and to extend XML with multimedia presentations, and investigate XML-embedded multimedia presentation storage, delivery, and querying. XML specifications do not yet incorporate content-based multimedia data modeling within XML documents. In addition, present semistructured data manipulation languages and data models do not yet deal with multimedia, much less with presentations and synchronization issues in XML documents in the web. This research will model content-based multimedia data as well as multimedia presentations within XML, revise/merge/consolidate XML language proposals and multimedia data/presentation languages, and investigate the query processing issues of the resulting newly defined languages.

Scope: The topic is widely recognized as a critical problem area for the next generation of global information systems. XML is clearly emerging as the standard according to which data (publications as well as other data) will be made available on the Internet. We are in the early stages of the full specification of XML and in the infancy of database support for XML data. The CWRU group has experience with multimedia databases. The group at Bilkent has complimentary expertise in distributed and real-time systems. Literature and other resources for the project are easily obtained from the Web; thus students in Turkey working on this project will not need specialized facilities. In addition to the collaboration with a colleague in Turkey, there will be training of a Turkish graduate student at the CWRU facilities. This project fits the mission of the Division of International Programs (INT) for support of mutually beneficial research.

Attributing function to genes, their regulatory elements and other functional DNA sequences will be the next step in genomics. This functional phase involves researchers from a variety of disciplines, including computer science and engineering. To facilitate the use of genomic informatin in function studies, pathways will be used to present the information. Pathways require detailed integration of diverse sources of information. Researchers will need transparent access that will be provided through this project. The project will integrate the data from a pathways rather than a sequence perspective, enabling scientists who do not do the sequencing to add functional information. New tools will enable users to visualize pathways data at mutiple levels of detail and abstraction. The use of XML will permit dissemination of query outputs for control of the data. Metabolic pathways, transcriptional regulation and protein synthesis pathways and signal transduction pathways will be the first incorporated in this web-based system. The research provides both for student support in both biology and computer science and a resource for the functional genomics community.

This project investigates web querying techniques for accessing web information resources. The term information resource refers to large web-accessible resources such as the ACM Digital Library. We propose a semantics-based way of accessing a web information resource: extract metadata about topics and relationships from the web resource, extend the metadata with "importance scores", and query it from a database. The query language is extended with constructs (a) to propagate importance scores to the query output to rank query output, and (b) to define "stopping conditions" to reduce query evaluation times.
For some query requests, the metadata in the database may not be sufficient to answer queries. Our research direction is to locate more informative query answers by mixing database querying with "focused crawling" in the web information resource, at the algebraic operator level of SQL queries. These queries allow time constraints, and relax the closed world assumption, making it necessary to redefine the notion of well-defined queries.
Data extraction techniques will be employed to extract metadata. Variations of our basic approach that do not require direct database query engine changes will be evaluated. Standalone web applications will be developed, and made available.

Abstract


Program: NSF 04-588 CISE Computing Research Infrastructure
Title: CRI: Infrastructure for Managing and Analyzing Large Scale Biological Data via Utility Computing
Proposal: CNS 0551603
PI: Yang, Jiong
Institution: Case Western Reserve University

Investigators at Case Western Reserve University will develop computing infrastructure to enable research in five bio-informatics application areas that have challenges in computing, data management and communications. Data intensive application projects include studies of biochemical pathways at multiple levels and searching and data mining in PubMed literature. Computation intensive applications include correlation discovery in two problem areas genetic variation in human populations and circadian rhythms. The communication intensive application is surgical planning simulators. The investigators will develop a utility computing infrastructure and explore computer science research issues such as load balancing, scalability, and multicast based data dissemination.

Proposals 0820217/0819987

Enhancing the Reliability of Bioinformatics Software by Correlating User Feedback and Execution Data

PIs: Andy Podgurski and Wassim Masri

A neglected consequence of the proliferation of scientific data sets and computational services on the Internet is that the scientific community is becoming increasingly dependent on the quality of shared data sets and the reliability of the software used to analyze them. To make it easier for developers of bioinformatics software to ensure its reliability, this research seeks to develop, evaluate, and refine automated techniques to help developers discover emergent reliability problems with deployed software, understand their nature and significance, and diagnose their causes. The approach is based on eliciting structured feedback from users about the problem symptoms observed and then automatically correlating this feedback with far more detailed information about internal program dynamics and input-output mappings. Advanced data mining techniques will be employed in tandem with dynamic program dependence analysis to: pool problem reports from different users; corroborate individual reports; group failures according to their symptoms and causes; and help diagnose the causes of failures. The proposed work has the potential to significantly improve the reliability of bioinformatics applications used by thousands of scientists.

Case Western Reserve University of Cleveland, Ohio is awarded to design and implement a database-enabled framework and tools to facilitate effective and efficient model development for multiscale mechanistic models of biological systems. The set of integrated database-enabled tools for regulatory metabolic networks, called PathCase-SB, will have enhanced functionalities to (a) link the ever expanding body of molecular information to an understanding of how intact organisms function via multiscale mechanistic models of the system, and (b) facilitate interactive model development and dynamic analysis of responses in an effective and efficient manner.

PathCase-SB will provide a framework and an environment that allows for the reproducibility, verifiability, and adaptability of the maintained models. It will have four interfaces, namely, Querying Interface, Visualization Interface, Simulation Interface, and Modeler?s Interface. PathCase-SB will be designed and built, to the extent possible, as a general-purpose web-based system for regulatory metabolic network models. The system will be built, tested and enhanced while carrying out (in silico parts of) a research on multiscale computational models of oxygen transport and metabolism in skeletal muscle. PathCase-SB model database will be integrated with PathCase, an existing metabolic pathways system, to enable building ?one-shop? querying and visualization capabilities.

The PathCase-SB project will provide a new and high impact visualization, querying, and simulation environment for regulatory metabolic network models. It will allow comparative analyses of results from different quantitative simulations, which will have the potential for improving the models by providing better insights on the suitability of the models. PathCase-SB will be an add-on, and complementary, to the metabolic pathways knowledge available in existing metabolic pathways data sources on the web such as Kegg, Reactome, or BioCyc. Finally, PathCase-SB will be suitable for educating students in systems biology, biology, bioinformatics, and medical sciences--to train them on the use of bioinformatics tools in terms of regulatory metabolic network models.

"This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5)."

Case Western Reserve University has been awarded a grant to build computational tools to facilitate research in metabolic profiles. Metabolites are molecules that are intermediates and products of metabolism. Current blood-labs in most hospitals and health care centers in the country can only measure in a blood sample small numbers of (less than 15) metabolites. However, recent advances in high-energy physics technologies has now made it possible (for researchers for the time being) to measure much larger numbers of metabolites in a blood sample, up to about 300 of them. When these measurements differ from those in a healthy individual (i.e., the control subject), the basic question is: what do they mean?

Currently, the above question can only be answered manually by experts in metabolic biochemistry. This project will test the validity of computationally and algorithmically deduced first-cut interpretations of such large sets of metabolite measurements. These interpretations are made available through web-based, database-enabled, easy-to-use, online and reliably correct software systems that are
- Scalable to the complete metabolic network, not parts of it and
- Grounded by the well-known and accepted metabolic biochemistry principles,
instead of being done manually as there are not enough metabolic biochemistry experts around to manually interpret each such lab result.

This project develops a web-based resource and its tools to enable the research community to help answer the above-listed question. The tools of the resource include
- Observed Metabolite Analysis tool that eliminates those metabolic paths that cannot have caused the observed measurements (i.e., inactive metabolic paths in the metabolic network), and provides a small list of activated metabolic paths that may have caused the observed measurements. Note that no effort is made to return actual flux estimations in the network, simplifying the task at hand,

- Metabolomics-Oriented Querying tool, that allows researchers/users to further focus on activated metabolic paths within a subnetwork (by allowing them to formulate queries about activated paths, futile cycles, etc.), and provides users with explanations of its query answers,

- Metabolomics-Oriented Visualization tool, which provides visualizations of activated paths within the complete metabolic network or zoomed-in and focused parts of the network, allowing users to visually interpret and analyze the results.

The resulting online tools will be an excellent web-based educational tool for educating students in metabolic biochemistry, biology, bioinformatics, and metabolomics--to train them on the use of bioinformatics tools in terms of metabolic network models. Additional information about the project may be found at http://nashua.case.edu/.