#216 Abstracts

ACS Chemical Information Division (CINF)
Fall, 1998 ACS National Meeting
Boston, MA (August 23-27)


Marriott Copley Place, Salon C, 4th Floor

Interfacing Robotics and Laboratory Automation
R. W. Snyder, Organizer, Presiding
9:00 1 Robotics, computers, and biochemistry: automation and the drug discovery process.
Marc N. Feiglin, Merck Research Labs, Rahway, NJ 07065.
Technological advances in the areas of Genetics, Combinatorial Chemistry, and High Throughput Biochemical Screening have revolutionized the drug discovery process. These technologies allow scientists to investigate more chemical entities against a larger number of biological targets in an increasingly shorter period of time. High Throughput Screening (HTS) is one area that has been tremendously affected by the introduction of computers and robotics. As automation continues to be implemented for drug screening, a typical HTS lab functions more like a factory and less like a traditional research lab. Automated systems are now involved at every stage of HTS at Merck including assay registration, sample storage, biochemical testing, data analysis, and data management. The integration of both commercial and custom in-house solutions to build a fully integrated system for HTS will be presented.
9:30 2 Advances in automation that make the sorting of 10,000 SPOS microreactors routine.
Anthony W. Czarnik, Vice President, Chemistry, IRORI, 11149 North Torrey Pines Road, La Jolla, CA 92037
The ability to synthesize small molecule libraries on the order of tens of thousands of members is becoming a standard job expectation in laboratories engaged in drug discovery. In addition to the obvious equipment requirements for miniaturization and chemical compatibility, the sheer numbers involved demands that reaction vessel handling become automated. Miniature radiofrequency memory tagged microreactors that are readable electronically have been developed to address this need. The movement of microreactors can occur at many stages of the split-and-pool library synthesis process. When thousands of microreactors are involved, the complete automation of these sorting requirements becomes desirable. In this talk, we will describe our reduction to practice of an automated sorter capable of moving thousands of microreactors to individual reaction vessels, removing this burden from the synthetic chemist.
10:00 3 Leveraging research informatics: bridging the automation information gap.
David Kniaz, 18 Campus Blvd., Newtown Square, PA 19073.
With the introduction of new technologies such as combinatorial chemistry, the number of samples being stored, tracked and retrieved has become a barrier for research. Automation can be used to exploit these compound libraries. This case study focuses on such a research facility that needed to upgrade the existing process of manually storing and retrieving its proprietary compounds in order to speed research and better manage its archive of research compounds. This was achieved through integration of a robotics system with data management software for the automated chemistry services. Automated sample management is now key to the efficient use of the vast proprietary compound libraries at the heart of high throughput screening and will provide scientists the freedom to perform the work they do best -- research.
10:30 4 Synthesizing robot programs from descriptions of combinatorial libraries.
David Chapman, Jonathan Handler, Afferent Systems, Inc., 2005 16th Street, San Francisco, CA 94103.
The difficulty of programming chemical synthesis robots is a significant barrier to their use. I will describe software (Afferent Synthesis) that eliminates the need for such programming, by working from an abstract description of the chemistry used to make a combinatorial library. The system is device-independent; with a suitable driver, it can be used with any synthetic instrument. Afferent Synthesis also makes it possible to interleave the synthesis of several libraries, or several batches of reaction vessels from a single large library, eliminating incubation "dead time" and increasing throughput. It provides robustness by implementing synthesis error recovery and task restart. A protocol optimization module allows the user to systematically vary conditions to find best conditions for a reaction. When coupled with other parts of the Afferent system, the software automatically generates chemical product structures, and tracks their locations.
11:00 5 Visual environments for collaborative drug discovery.
Christopher Ahlberg, Spotfire, Inc., 28 State Street, Suite 1100, Boston, MA 02109
The data explosion caused by modern laboratory equipment causes a fundamental problem in how researchers are to be able to cope with the flow of data. Visual environments for data exploration promises the ability for researchers to rapidly finding trends and patterns, identifying anomalies and detecting often crucial gross early errors. By taking advantage of modern graphical desktop computing power, users can perform orders of magnitude many more visual tests of data sanity - a strong advantage for example in high throughput screening and micro array analysis. The presentation will include a live demonstration of how visual discovery technology can enhance a LIMS system.
11:30   Lunch Break
2:00 6 Synthesis on a chip.
Sheila H. DeWitt[1], William Chiang[2], Yue Chen[1], Timothy Hamilton[1], Sean Kelley[1], Jack Leber[3], Sterling McBride[2], Richard Moroney[2], Kerry O'Mara[2],
[1]Orchid Biocomputer, Princeton, NJ 08543
[2]Sarnoff Corporation, Princeton, NJ 08543
[3]SmithKline Beecham, King of Prussia, PA 19406.

A microfluidic, chip-based system has been developed and used for the solid phase organic synthesis of discrete compounds in a massively parallel array. Up to 50 nmol of a single compound can be generated in each 700 nl reaction well. The use of electrohydrodynamic pumping enables the transfer of organic solutions that have little or no conductive nature, eliminates the need for mechanical valves, and provides the maximum diversity of reagent delivery for optimally designed libraries. The synthesis of up to 100 compounds on one chip employing 1-3 reaction steps has been achieved using electronic and/or pressure pumping. The results of these syntheses and issues related to the transport of solutions, reagent monitoring, and quality control of products will be presented.
2:30 7 Reagent selector: a new tool for high throughput synthesis.
Maurizio Bronzetti, Al J. Gushurst, Doug R. Henry, and Robert W. Snyder, MDL Information Systems, Inc., 14600 Catalina Street, San Leandro, CA 94577.
Reagent selection is one of the most critical and strategic steps in synthesis planning and library design. Availability, synthetic feasibility, structural or property similarity/dissimilarity and cost are some of the criteria that chemists use today to select reagents for a specific library synthesis. These criteria often determine the success or failure of a lead generation or optimization program. Reagent Selector is a new software tool for reagent selection. It is a bench tool created for combinatorial, medicinal, and synthetic chemists who often spend hours filtering and classifying their in-house or commercial reagent lists for synthesis. With Reagent Selector, scientists can also manage personal and sharable lists of reagents for subsequent synthesis work, create pick-lists for collection purposes, and generate purchasing lists for ordering. The program features property calculation and clustering capabilities which can be customized and expanded by the user.
3:00 8 Integration of informatics with a robotics synthesis laboratory.
K.I. Heuer, D. L. Flynn, J. F. Gaw, Searle Discovery Pharma IT, Monsanto Company, 4901 Searle Parkway, Skokie, IL 60077.
Combinatorial chemistry has aggressively introduced automation into the Discovery process. The use of these robotics automation tools has produced a wealth of discovery data. This presentation will describe the software tools and procedures developed to interface the chemistry and the chemist to the robotics synthesis laboratory, and manage the massive amount of data generated in the automated discovery process. Highlighted will be our Electronic Spreadsheet tool with which we capture robotics and manual synthesis steps, analytical interpretations and product yields. Also noted will be our SARplus tool which we developed to schedule different pathways of biological screens. This integration provides chemistry throughput and intelligent discovery data management.
3:30 9 Automating data analysis for high throughput screening.
Susan I. Bassett, John W. Elling, Bioreason, Inc., 309 Johnson Street, Santa Fe, NM 87501. g
High throughput screening yields more positive responses (hits) than can easily be analyzed manually. In addition, information is contained in the large number of negative responses from inactive molecules. Bioreason has developed a suite of automated methods that aid chemists in lead identification and optimization, using structural and physical feature information both from the hits and inactives. In our systems we use a variety of computational intelligence techniques, including traditional methods such as expert systems and statistical pattern recognition as well as neural networks and fuzzy reasoning systems. In this talk, we will present a system to automate selecting training sets directly from the screening data to be used in pharmacophore model building and QSAR analysis.
4:00 10 High throughput: the hidden bottlenecks.
Ron Niesen, 18 Campus Blvd., Newtown Square, PA 19073.
The frenzy is on. Any and all research organizations are putting every ounce of their efforts into building a "discovery factory." But are they destined to repeat the same learning curve as manufacturers did over the past century, or will they incorporate the best innovations of modern manufacturing into their factory approach? With the "discovery factory" as an analogy this talk will discuss the characteristics of bottlenecks, both from a technical and an organizational viewpoint. It will highlight strategies for addressing hidden ones, particularly regarding information, that can provide significant improvements to process throughput. Setting up a high throughput factory requires more than piecing equipment together. The audience will leave with a an understanding of where to focus their efforts for the most results using leading and available technologies.



Marriott Copley Place Salon C, 4th Floor

TriSociety Symposium
B. Slutsky, Organizer, Presiding


Introductory Remarks
Bruce Slutsky


How the Web may change the way science librarians are trained in graduate schools of library and information science.
Tony Stankus, Science Library, College of the Holy Cross, Worcester, MA 01610-2395
A number of factors have deterred the ready training of future academic librarians proficient in the information needs and sources in the sciences in many library/info science schools in the U.S. These include the overwhelming preponderance among future librarians of science-limited undergraduate backgrounds in English, history & education, most of whom are second-career adult learners, their increasingly part-time, and distance commuter status, and the uncertain junction of strong collections in the sciences on the same campus as the library school. The web provides substantial possibilities for confidence building exercises for non-science students, reduces the dependency on sometimes limited local print collections, and builds navigational skills via remote access to web-based databases and electronic journals. Developing these skills also has crossover potential in the world of corporate librarianship.


The CINF Web site.
Andrea B. Twiss-Brooks, The John Crerar Library, University of Chicago, Chicago, Illinois, 60637
The Division of Chemical Information of the American Chemical Society is an organization with the mission of "...providing a forum for the exchange of information and expertise among the generators, developers, providers, and users of chemical information worldwide..." In 1995, the Division of Chemical Information (known by the acronym "CINF") undertook to expand its efforts to promote the exchange of information by beginning the construction of a web site. The site has grown from a two small HTML documents to a multiple level presence on the web, providing information about CINF technical symposia, conferences, workshops, membership, divisional reports, and more. The history, growth and development of the web site will be presented, as well as statistics on the use of the site.
9:30 13 ChemCenter: The ACS guide to navigating chemistry online.
Louise Voress, Sarah W. Nash, and David A. Koran, American Chemical Society, 1155 16th St., NW, Washington, DC 20036
This paper gives a summary of Internet resources currently available on ChemCenter, the Web site for chemistry professionals (including ACS members and nonmembers) in industry, academe, and government, as well as a resource for educators, students, and the general public who want information about chemistry. This "virtual society" is a central source for information and interaction with ACS, its programs, and activities. It affords users the opportunity to learn about and link to other Web sites, such as ChemPort, the research site service of ACS Publications and CAS; highlights information of importance to its audience from ACS and other credible sources; offers unique features intended to encourage users to return to the site on a frequent basis; helps users organize the vast resources available electronically; and serves as a reliable Web starting point that can be accessed internationally, 24 hours a day.
10:00 14 Weaving the Web into the CAS product line.
Jan Williams, CAS, Columbus, OH 43210
Bringing chemical information products to the World Wide Web is a natural step in the evolution of the CAS product line which began with print in 1907 and moved into electronic delivery with online and CD delivery in the eighties and nineties. The new members of the family -- Chemical Patents Plus, STN Easy, and the Chemport connection to full text journals -- offer both information professionals and scientists different capabilities for access to information as well as increased opportunities to integrate the rich content of Chemical Abstracts into their work. However, these web products complement rather than replace the CAS hard copy and electronic services. This paper defines the web products and positions them within the broader product line in terms of potential users and applications.


The ISI Web of science.
Matthew Clark, Institute for Scientific Information, 3501 Market Street, Philadelphia, PA 19104
The interlinking nature of the Web makes it an ideal medium for scientific information exchange. ISI has created a powerful information resource by using the Web to link chemical reactions, articles, cited references and citations, document delivery services, and publishers' full text. The ISI Chemistry Server allows substructure and reaction searching through a web browser. Chemists can search on transformations, reaction conditions, and catalysts. Then scientists can see the references for an article, who has cited it, click on these items to navigate to those articles, and link to the full- text.
11:00 16 ACD/ILab as a universal Web front-end to chemical information resources
Valeri Kulkov, Antony Williams, Advanced Chemistry Development, Inc., 133 Richmond Street West, Suite 605, Toronto, Ontaria M5H 2L3, CANADA
The Interactive Laboratory, ACD/ILab, is a universal Web front-end to chemical information resources, property prediction programs and chemical databases. ILab utilizes Java-based structure drawing and spectral display applets to provide structure submissions and spectra visualizations. We have developed an open server interface for connecting third party chemical information resources to the ILab server. The actual resource can be located anywhere on the network, thus enabling a uniform user interface to chemical information that is distributed globally. Currently, the following database searches and property predictions are available at ILab: IUPAC and CAS name generation, H-1 NMR and C-13 NMR spectrum prediction, pKa database search (over 9000 structures), pKa prediction, LogP/LogD prediction, BP prediction, Vapor pressure prediction, etc. and others, 22 services in total.
11:30   Lunch Break
2:00 17 The NIST WebBook - A simple tool for chemical data access.
W. Gary Mallard, Peter J. Linstron, NIST, Gaithersburg, MD, 20899
The NIST WebBook is designed to make access to all chemical data, but especially evaluated data from NIST available with minimal effort. The design goals of the WebBook are simplicity and ease of use. All data are given full references so that the original sources may be examined. Where equations are provided JAVA applets are provided to graph and blow up the data. Spectroscopic data is also distributed via JAVA applets so that the ability to blow up sections of a spectrum are present. A discussion of the design, coverage, utilization and future of the WebBook will be included.
2:30 18 Providing links between secondary and primary electronic literature.
S.J. Young, W.D. Hounshell, M.A. Duda, E.V. Gordeeva, MDL Information Systems, Inc., 14600 Catalina Street, San Leandro, CA, 94577
The ability to directly search electronic literature is incredibly powerful. For example, the indexes to primary electronic literature enable one to search any word in any journal article. One new common access method to the primary electronic literature is from commercial molecular or reaction databases. The development of such links from structure databases to the primary literature will be discussed. Examples including linking MDL's reaction databases to Science Direct, Elsevier Science's electronic journal subscription service, will be presented.
3:00 19 Do chemists need clubs or communities on the Web?
Wendy A. Warr, Wendy Warr & Associates, 6 Berwick Court, Holmes Chapel, Cheshire CW4 7HZ, England
Virtual communities have their origins in early PC bulletin boards but the Internet has enormously increased their significance not just for users, but for enterpreneurs, investors, publishers, and Web developers. Not all newsgroups and discussion forums are necessarily virtual communities. A virtual community is distinguished not just by a common interest but by "membership", commitment by the members, and interactivity. This paper will compare some emerging virtual communities in chemistry, in particular ChemCenter and ChemWeb.com, and the facilities they offer (news, job centers, shopping malls, libraries, databases, etc.). The paper will analyze whether chemists do form one distinct community and whether the services on offer are appealing and usable in their present form.
3:30 20 ChemWeb - The world wide club for the chemical community.
William G.Town, , ChemWeb, Inc., 34-42 Cleveland Street, London, England, W1P 6LB
ChemWeb is a web community for chemists which offers to its members a library of electronic journals, text and structure searchable databases, a webzine (The Alchemist), a job exchange, a conference diary, a shopping mall and experimental services including live virtual conferences. Also free to members of ChemWeb is ChemDex Plus, a searchable and browseable directory of chemical web resources. As ChemWeb evolves, empirical lessons are being learnt about electronic communities in general and the needs of chemists in particular. The possible future development of such a community will also be presented.
4:00   CINF Business Meeting
4:15   Intermission
4:30   Concurrent Open Meetings for Society Committees on Publications and CAS



Convention Center Exhibit Hall B, Plaza Level

C. Gragg, Presiding
7:00 - 9:00
  22 Dental patent literature
Peter F. Kulkosky, U. S. Patent & Trademark Office, Arlington, VA 22202
The U. S. Patent & Trademark Office class and subclass categories are definitive of collections of patents relating to every area of dental practice. Searching of same will be explained with emphasis upon the use of the Automated Patent System (APS). International Patent Classification (IPC) and commercial database scanning will be compared to APS. Sample use of APS search commands specific to the Title, Abstract, Drawings, Figures, Claims and other sections of U. S. patents in the dental fields will be illustrated.
  21 Spatial perception of real vs. unreal molecular models using virtual 3-depth enhancement.
Bert Ramsay, Chemical Concepts Corporation, 912 N. Main St., Ann Arbor, MI 48104, William J. O'Neill, Colleen O'Neill, D/E Technologies, Inc., 107 Aprill Drive, Suite 2, Ann Arbor, MI 48103, W. V. Metanomski, Chemical Abstracts Service, 2540 Olentangy River Road, Columbus, OH 43210.
Recent studies have suggested that the development of spatial perception skills is important in the understanding of chemistry concepts. ( for example, see S. L. Coleman and A. J. Gotch, J. Chem. Ed., vol. 75 (2), 206 - 209, 1998). It is not clear whether the increasing use of molecular modeling programs will allow chemists to distinguish between real and unreal molecules. Although the connection tables for cubane and e-cubane are identical it is sometimes difficult to see which one is an illusion. We believe that the spatial perception of a 2-D molecule can be enhanced using the Virtual Window © technology that does not require the creation of a 3-D image. Come see it - to believe it!



Marriott Copley Place Salon C, 4th Floor,

Skolnik Award Symposium on Electronic Publishing
G. D. Wiggins, Organizer, Presiding
8:30   Introductory Remarks
8:45 23 Award Address - A PEAK at 25 years of an academic chemistry librarian's career (in B Major).
G. D. Wiggins, Chemistry Library, Indiana University, Bloomington, IN 47405.
The Indiana University Chemistry Library has long been a leader in the application of new technology to chemists' information needs. From the early days of spinning Chemical Abstracts tapes in-house to the current provision of access to a wide array of Web-based and in-house databases spans nearly three decades. An overview of these developments will be presented, culminating in the most recent experience of providing information about chemical information sources through CHMINF-L (the Chemical Information Sources Discussion List) and experiments with electronic journal access (including, the Elsevier PEAK project: Pricing Electronic Access to Knowledge).
9:30 24 New methods of communication amongst chemists.
Wendy A. Warr, Wendy Warr & Associates, 6 Berwick Court, Holmes Chapel, Cheshire, CW4 7HZ, England.
This aurthor's business revolves around networking, competitive intelligence and "primary research", that is, finding information that is so current that it has not yet, or never will, appear in patents, journals, databases, etc. The "invisible college" exchanges information in person, over the telephone and at conferences. The possibilities for computer-mediated communication, and not just electronic mail, now add a new dimension. Listservers are one option but many of these are used largely by "lurkers", that is, by listeners rather than contributors. Some electronic journals allow discussion on papers and topical issues. Electronic conferences in chemistry have made some impact over the last year or two but few of them are real-time, with the advantages and disadvantages entailed. This paper will discuss some of the cultural, social and technical issues in computer-mediated communication amongst chemists.
10:00 25 Current issues in electronic publishing: the HighWire Press perspective.
Michael Newman, Head Librarian and Bibliographer, Falconer Biology Library, Stanford University, Stanford, California 94305-5020.
The HighWire Press team at Stanford University was assembled in 1995 to develop the World Wide Web version of the Journal of Biological Chemistry. Since that first project, HighWire has developed Web versions of dozens of additional high-impact biomedical journals, implemented access control for subscriber-only access to some titles, implemented inter- journal links and cross-journal searching, and introduced a range of new features that take advantage of the electronic medium. As a unit within an academic library working with scientific publishers, High Wire occupies a unique position in the interchange between publishers and information consumers. This paper will focus on some of the current issues in electronic publishing viewed from this unique HighWire perspective. One of these issues is the need for pricing models for electronic journals that promote the flow of published information while preserving the publisher's revenue stream. Related to pricing models is the need for an archival medium for electronic information. Finally, the Web offers new opportunities to link related information. HighWire and other providers of electronic information already offers a variety of links, and new links will provide opportunities for cooperation among publishers to enhance the delivery of information to users.
10:30 26 Emerging trends in electronic full text delivery
Eileen M. Shanbrom, Chemical Abstracts Service, Columbus, Ohio 43210.
With the demand for increased productivity in scientific research today, scientists need to "close the information loop" by obtaining quickly the full text of articles key to their research. In the past, scientists have left searching information and obtaining articles to information professionals and librarians who have relied on secondary publishers and document delivery services. With new technology today, scientists are searching for and acquiring information on their own. Ordinary Web search engines are not adequate for effective searching of scientific literature. Innovative cooperative arrangements between secondary and primary publishers are necessary to efficiently deliver the goods. Traditional document delivery suppliers, individual publishers, subscription services, and information providers are offering many choices for obtaining documents. This paper discusses trends and possibilities for rapid delivery of full-text information. Several new product developments at CAS are described, including a new model for aggregation, ChemPort.
11:00 27 Dream on: visions of academic access in an era of integrated chemical information systems.
Ken Rouse, 2361 Chemistry Building, University of Wisconsin-Madison, 1101 University Avenue, Madison, WI 53706, and Katherine R. Porter, Chemistry Library, PO Box 90355, Duke University, Durham, NC 27708-0355.
As the pace of the electronic revolution accelerates, academic chemistry librarians are excited by prospects of an integrated approach to chemical information which will soon enable end-users to move effortlessly between powerful electronic indexes and fully searchable, full-text journal databases. Unfortunately, the pricing paradigms proposed thus far by many information providers -- whether journal publishers of creators of indexes -- put this tantalizing electronic future beyond the reach of even some of the largest universities. Welcome to the age of the "information rich and the information poor."
12:00 28 CINF Luncheon: "A two-hundred-year perspective on CINF on the occasion of its fiftieth birthday."
Mary E. Bowden, Leo B. Slater, The Chemical Heritage Foundation, 315 Chestnut Street, Philadelphia, PA 19106-2702.
This paper will consist of a brief illustrated tour around the Chronology of Information Science prepared by the Chemical Heritage Foundation to celebrate the fiftieth anniversary of CINF. Major eras and outstanding developments by pioneers in chemical information science will be highlighted--from the founding of the first chemistry journal in 1778 and the nomenclature reform proposed by Lavoisier and his followers in 1787 to the present. As will be shown, CINF's founding in 1948 occurred at a critical nexus in chemical information science in terms of the volume of information being produced, new technologies available to master this flood, government interest in supporting such efforst, and in the very definition and organization of an emerging profession.
2:00 29 Developing E-collections: It's not the years, it's the internet mileage.
Kimberly J. Parker, Yale University Library, 130 Wall Street, P. O. Box 208240, New Haven, Connecticutt 06520-8240.
Collecting electronic resources can seem intimidating at first glance. A selector should consider presentation and training factors, access control methods, licensing terms, technical and systems related issues, pricing models, usage statistics, duplication in different formats, consortial decisions, archiving and ownership issues, and impacts on other library units. Broken down into parts and examined as an extension of selecting and acquisitions practices in the non-electronic world, the process becomes easier. Selectors do not require years of experience in economics, programming, copyright law, or design work to collect materials in electronic format. They simply need to have an email network of similarly minded colleagues and access to Internet web-sites.
2:30 30 Electronic journals in the academic library: issues of copyright and plagiarism.
Veronica Calderhead, John Cotton Dana Library, Rutgers, The State University of New Jersey, Newark, New Jersey 07102.
Electronic journals and the easy access to them in academic libraries has ushered in some unanticipated problems faster than libraries have been able to deal with the issues. Copyright, i.e., the legal and contractual issues surrounding the fair use of the ubiquitous journal, is complicated and understood by a handful of individuals at a given institution, even though the journals are used by hundreds of users. Plagiarism is an ethical issue typically detailed in the honor or behavior codes of an academic institution. These codes were written long before the Web and electronic journals became part of the university libraries' collections. Both of these issues, copyright and plagiarism, will be examined in the context of the research library. Preliminary data will be presented that indicates a disturbing trend in the unethical use of electronic articles.
3:00 31 Doctoral dissertations graduate to the Web: new applications and implications for scholarly research.
B. Lawlor, Chescot Publishing, Inc., P. O. Box 556, Wayne, PA 19087.
Beginning in 1938, the publishing of dissertations became an integral requirement for obtaining a doctoral degree in the United States. Until recently, dissertations were published exclusively in print and/or microform, copyright issues were similar to those of any print material, and, for the most part, any given dissertation was not widely accessed or read. With the advent of the World Wide Web, doctoral dissertations have taken on a new importance in the area of technology transfer from academia to industry and they are even more valuable as part of a University's intellectual property. But electronic dissertations create new problems with copyright, archiving, and, from a student's perspective, in the ability to get doctoral research published in prestigious print journals. These and other issues resulting from " Dissertations -- Graduating to the Web" will be explored.
3:30 32 Technical challenges in producing ACS journals on the World Wide Web.
Lorrin R. Garson, Advanced Technology Department, Publications Division, American Chemical Society, 1155 Sixteenth Street, N. W., Washington, DC 20036-4892.
The American Chemical Society began making its primary journals available on the World Wide Web in the summer of 1996 with all of its 26 titles being available in September of 1997. The salient features and functions of the Web journals will be discussed and an overview of the production technology will be described. Other related topics that will be discussed are: (a) the relationships between databases and print and electronic production, (c) economics of electronic journals, (d) the concept of ASAP (As Soon As Publishable), and (e) the impact of standards on electronic journal creation.
4:00 33 The complete chemical E-publication.
Peter Murray-Rust, University of Nottingham, Nottingham, NG7 2RD, UK.
The development of new tools for use on the WWW is now extremely rapid. Even allowing for the current "hype" over XML (eXtensible Markup Language) and other protocols, it seems certain that most Web-based information systems will be adopting them. The announcement by major suppliers that they will be developing XML-based browsers and editors means that there will be a large number of affordable high-quality tools available very shortly. The goal is to make information available globally, in any discipline, and as easily as possible to authors and readers/users. Authors and tool developers will use documents and data from different domains that interoperate in a platform- and vendor-independent manner. The XML family of protocols will allow integrated documents and data for the first time. To support chemistry, we need to address specifically molecular problems. Although there are no agreed semantics for molecular data nor de facto standards, a starting point, Chemical Markup Language (CML) will be presented.
4:30 34 Databases - the journals for the new millenium.
Stephen R, Heller, NIST, Gaithersburg, MD 20899.
This talk will describe the author's view of what much of scientific publishing in chemistry is likely to look like in 20-30 years. As chemists are more interested in a need more actual factual data (chemical and physical properties, chemical structures, reactions, and so on), the ability to easily submit, retrieve, use and correct data in a database will prove much more attractive and valuable than the current model of publishing a scholarly manuscript as a separate and independent entity.
5:00   Concluding Remarks
Gary Wiggins, Skolnik Awardee



Section A
Marriott Copley Place, Salon C, 4th Floor


Competitive Intelligence: Sources and Applications
T. Wright, Organizer, Presiding
8:30 35 Collecting human intelligence.
J. Michael French, Anthony M. Page, HealthIQ, 750 The City Drive, Orange, CA 92868.
To support effective decision making, organizations must learn to obtain the best information rather than relying on the most accessible. While effectively using secondary sources and internal contacts is necessary, this is only the starting point for a professional collection effort, which includes information from human sources (HUMINT). Obtaining HUMINT often requires skillfully conducted primary interviews. This type of collection can be done legally and ethically and should not be confused with commercial "espionage", although legal issues must be clearly understood to minimize the risk of improprieties. A clear understanding of the unique aspects of competitive intelligence collection is also necessary in order to maximize an individual's ability to obtain useful information. These aspects include knowing how to: identify and assess potential sources; develop an effective approach; measure the reliability and validity of information; debrief collectors; and document interviews so that information can be efficiently integrated into analysis.
9:00 36 Competitive Intelligence gathering and analysis in biotechnology.
Aris Persidis, Argonex, Inc., 2044 India Road, Suite 202, Charlottesville, VA 22901.
Biotechnology is information rich, both in terms of scientific and also business developments, whose pace and progress is dramatic. Successful participation in the field requires a systematic approach to information gathering and analysis. Competitive intelligence is defined as actionable information, which requires the ability to filter and synthesize relevant knowledge. Various approaches to competitive intelligence gathering and analysis will be discussed, both from the user but also from the intelligence provider perspectives. Examples will be drawn from biotechnology, but the approaches are applicable to all high-tech fields.
9:30 37 Providing effective early warning.
Anthony Page, HealthIQ, 750 The City Drive, Orange, CA 92868.
The potential for surprise is inherent in any decision regarding future operations because the future is fundamentally unknowable. Organizations try to cope with this dilemma by basing their decisions on well-reasoned intelligence estimates which forecast the future by extrapolating from current information. Any such estimate, however, is limited by the accuracy of its assumptions which inevitably diverge from reality with the passage of time. Occasionally these assumptions diverge significantly from reality, serious problems arise as a result, and organizations then wonder how they were "surprised". The Indications and Warnings (I&W) Intelligence methodology addresses this problem by enabling the intelligence manager to construct an early warning system. This early warning system addresses the fundamental human and organizational obstacles which normally prevent us from perceiving significant change and reacting appropriately. First, it provides a technique for systematically tracking key assumptions into the future. Second, it provides a mechanism for providing timely and unambiguous warning to decision makers. In short, I&W intelligence is a control system which allows the decision maker to change course in reaction to measurable changes in the external environment.
10:00 38 Use of commercial information sources in the pharmaceutical area for competitor intelligence.
Alexander Mullen, Martin Blunck, Klaus Eike Moeller, Hans-Georg Rohbeck, Bayer AG, Phramaceutical Research Center, D-42096, Wuppertal, Germany.
Competitor information is available in electronic form via a variety of commercial producers covering the whole spectrum - from patenting activities, scientific literature, pipeline products, pharmaceutical industry news feeds, and broker studies to information about launched products and discontinued projects. These sources can be used, for example, to analyze and assess the strengths and weaknesses of competitor organizations, developments/ trends in specific therapeutic areas, etc. However, analyses based on such data are full of pitfalls for the unwary - as critical questions must first be addressed relating to their reliability, currency, and actual scope
10:30 39 Patent analysis in the pharmaceutical industry.
Robert Cantrell, Derwent, Inc., 1725 Duke Street, Alexandria, VA 22314.
Patent intelligence in the pharmaceutical industry is focused in three major areas: the inception of an idea and project, the synthesis and screening of compounds, and the application for patents of promising compounds. The first case involves screening for any company already claiming a compound's discovery and screening for other approaches to solving a problem that do not involve the proposed compound. The second case is an ongoing current awareness of research teams known or discovered to be competing in the same field. The third involves a last check for patentability and the decision as to how to protect the compound from a global perspective. Within these three phases, a company will analyze its current patent portfolio, those of its competitors, make forecasts regarding the future of respective portfolios, and examine pharmaceutical markets to better assess where research should be focused. For the most effective use of patent information, primary patent intelligence activities conducted by research and legal teams will be combined with business scenario planning and portfolio analysis.
11:00 40 Using federal trademark applications for competitive intelligence.
Karen Luszcz, The Dialog Corporation, 28 State Street, 23rd Floor, Boston, MA 02109.
Federal Trademark applications can provide a wealth of information in terms of tracking the product development activities in which your competitors are engaged. This session will focus on the lifecycle of a pharmaceutical trademark application, how this information can be used competitively and tips on how to secure a great name for a product under development.



Section B
Marriott Copley Place, Provincetown/Orleans, 4th Floor

Materials Science Informatics:Databases for Materials Design
J. Rumble, Organizer, Presiding
8:30 41 An integrated database approach to materials design.
John Rumble, Jr., Standard Reference Data Program, National Institute of Standards and Technology, Gaithersburg, MD 20899
During the past two decades, the advent of modern structural databases have fueled an unprecedented era of discovery in the biomedical sciences. New pharmaceuticals have be designed, biologial activity elucidated at the molecular level, and genetic coding disentangled. Today, non-organic materials science is ready for the same advances. Robust improved databases of structural information for metals, alloys and inorganic compounds are under development. These databases, combined with databases of phase equilbria and basic physical properties, hold the promise of new discoveries in surface science, catalysis, high temperature superconductors, ceramics, and alloys. This talk will review efforts to identify needed databases and how groups around the world are working to integrate together related databases so that scientists can derive the greatest benefits. Specific examples will be given from the NIST Standard Reference Data Program.
9:00 42 Crystallographic data for non-organic materials.
Vicky L. Karen, National Institute of Standards and Technology, Gaithersburg, Maryland 20899.
Crystallographic databases have traditionally been used as an aid to scientific research through bibliographic searching, the identification of unknown materials, and the analysis of individual crystal structures. The materials design community requires computational means to analyze, cross-correlate, and examine relationships among all data types, and is placing new demands on crystallographic data activities. To help meet these needs, NIST has extended its crystallographic data program to provide evaluated full structural data for all non-organic materials including inorganics, metals and minerals. Forthcoming products include the delivery of data items in standardized formats, the characterization of materials based on chemistry search modules, and three-dimensional visualization of structures. In addition to search, retrieval and display tools, scientific algorithms are being developed to assist information synthesis and knowledge discovery.
9:30 43 Surface structure data and materials design.
Philip R. Watson, Dept. of Chemistry, Oregon State University, Corvallis, OR 97331, Michel A. Van Hove, Lawrence Berkeley Lab., Berkeley, CA 94720 and Klaus Hermann, Fritz- Haber Institut, D-14195 Berlin, Germany.
The design of new materials for electronic, structural, corrosion inhibition, catalysis and other applications often hinges on a knowledge of surface, rather than bulk properties. One of the most fundamental items of surface information is a knowledge of the surface structure of a material - that is the crystallographic relationship of atoms in the surface layers of the substrate material itself and any adsorbed layers. The NIST Surface Structure Database (SSD) contains critically evaluated data for all published surface structures that contain hard information on interatomic distances and angles. Included is data on the experimental procedures, theory, 2D unit cells, 3D atomic coordinates, bond lengths and bond angles Moreover, structures can be interactively viewed and interrogated in 3D and printed/saved for later analysis. The latest version (SSD 3 .0 for PC/WIndows, available late 1998) will be demonstrated.
10:00 44 Computing macroscopic properties from microstructures of materials.
W. Craig Carter, Dept. of Materials Science and Engineering, Massacheusetts Institute of technology, Cambridge, MA 02139, Stephen A. Langer, NIST, Gaithersburg, MD 20899
The macroscopic properties of materials depend both on the properties of constituent materials as well as their microstructure. Microstructure involves the complex spatial arrangement of polycrystals, second phases and defects. The complex nature of microstructural data makes direct calculation of macroscopic properties a difficult task; nevertheless, many practical applications of materials depend on their precise microstructure. We have created public domain software which operates from scanned images of microstructures and allows the superposition of material properties directly onto the image. Virtual (computational) tests are then performed on the microstructural data and effective properties can be determined. Examples will be shown where thermoelastic properties of microstructures are calculated as well as reliability calculations.
10:30 45 Databases for advanced ceramics design and use.
Stanley Dapkunas, Ceramics Division, National Institute of Standards and Technology, Gaithersburg, MD 20899.
Excellent thermal, physical, electrical, magnetic and strength properties, good durability, and new processing methods have combined to make advanced ceramics a focal point for new engineering materials. In many cases, obtaining the desired properties for a specific material is dependent on seemingly small changes in processing, composition, and/or structure, e.g., the critical temperature of new high temperature superconductors is strongly dependent of small changes in composition. NIST is developed a suite of databases to support the design and use of advanced ceramics. The databases include data on structural and thermal properties, phase diagrams, high temperature superconductors, machining, and triboceramics. These databases and others are being integrated into a single system as the NIST Ceramics Webbook. The Ceramics Webbook also contain tools to facilitate analysis and manipulation of the data. This talk discusses the impact of the Ceramics Webbook on materials design. (http://www.ceramics.nist.gov/webbook/webbook.htm)
11:00 46 Materials databases for advanced performance.
David Rose, AMPTIAC, Rome, NY, 13440-6920
The Advanced Materials and Processes Technology Information Analysis Center (AMPTIAC) is one of thirteen Information Analysis Centers (IACs) funded by the Defense Technical Information Center (DTIC). AMPTIAC is responsible for maintaining the IAC's programs on metals, alloys, ceramics and composites. The demand for higher materials performance is driving efforts to consolidate and increase AMPTIAC's work in collecting and evaluating material properties. Today special attention is being paid to data on metal alloys for marine applications. Complementing the data collection and evaluation work is the development of an expert system to support materials selection for marine structures and components. Ultimately AMPTIAC plans to develop a family of property databases for many materials classes to support broad usage in advanced applications. The databases will be described in some detail with emphasis on those supporting the design of new materials.



Marriott Copley Place, Salon C, 4th Floor

Chemical Information on Intranets
A. Berks, Organizer, Presiding
2:00 47 Developing chemistry-enabled Web applications using Accord.
G. A. Hopkinson, A. P. F. Cook, J. G. Littler, Synopsys Scientific Systems Ltd., 175 Woodhouse Lane, Leeds, West Yorkshire, LS2 3AR, United Kingdom.
Publishing chemical information on the World Wide Web or creating chemically- aware intranet applications requires a means of phrasing chemical queries and visualising chemical moieties on the client, performing chemical searches on the server and transferring data to and from other desktop applications, such as chemical spreadsheets, databases and editors. Accord chemical components have been developed to provide solutions to these problems. The Accord ActiveX Chemistry Control is a client-side component that understands a variety of chemical formats, can create publication-quality chemical diagrams, and supports double-click editing using OLE-enabled chemical drawing packages. Accord/Universal Server, Synopsys' next generation chemical-relational database solution that allows chemical databases to be searched via the Web, will also be demonstrated.
2:30 48 The challenge of using extranets to improve data accessibility - the Derwent perspective.
Jan Kuras, Tim Miller, Derwent Information, 14 Great Queen Street, London WC2B 5DF, UK.
For the pharmaceutical industry to accelerate drug discovery and development, scientific information must be made more accessible to end-users. Information providers such as Derwent are meeting the challenge by presenting data using Web technology. By using "Extranets", data is reaching users more effectively, and data maintenance requirements are reduced - thus giving Information Specialists more time to carry out complex data analysis. It is also providing greater opportunity to share information throughout companies thus enahncing the knowledge management process. Derwent is making two of its premier biomedical databases, World Drug Alerts and the Derwent Drug File, available via Extranet. The aim of our presentation is to share our development experiences as we work with our customers to enhance data presentation, functionality and accessibility.
3:00 49 ColbyChem: a Web server for ISIS/Host.
John W. Kuehne, Thomas W. Shattuck, Colby College, Waterville, Maine 04901.
ColbyChem is a web-based server for MDL's ISIS/Host, built using software from the Apache/Perl Integration Project. ColbyChem brings most of the functionality of ISIS/Host to the browser by providing a dedicated web-server and ISIS/Host session for each user. ColbyChem permits searching of molecular and reaction databases, maintains hit lists, and displays search results. Essentially no configuration is required: data are returned in the natural hierarchy of the database determined by the hview and an optional fields file. This novel use of the Apache/Perl server can be generalized to other tasks. The browser becomes the display server for a remote application written in perl, with one process for each user, reminiscent of an X server connected to an X client. In contrast, the intended use of a web server is to transmit documents efficiently to many anonymous users.
3:30 50 Turning information into innovation: 3M Information Services Web site.
John D. Arenivar, Carol E. Herzberg, Sharon M. Peterson, 3M Information Services Patent Services, 3M Center Building 201-2C-12, St. Paul, MN 55144-1000.
As Web sources and services proliferate, corporations want to find efficient, effective ways to tap in to the Internet and turn what's "out there" into competitive advantage. 3M Information Services has developed an Intranet web site that links to evaluated free and premium Web sites, backing it up with education programs and support services. Patent information is a special focus, featuring several types of search and delivery options for the end user.
4:00 51 Web-based productivity tools for chemists: WebLab(TM) MedChem and Diversity Explorer.
Robert D Brown, Osman F Güner, Matt Hahn, Hong Li, MSI, San Diego
Computational tools are playing an increasingly important role in the decision making processes of combinatorial and medicinal chemists. They aid in managing the vast information flow resulting from high-througput medicinal chemistry, help select compound sets for lead finding and facilitate the rational decision making involved in lead optimzation and drug design. Web technology provides this opportunity. Pharmaceutical companies are effectively using their corporate intranet to communicate and retrieve information. Web browsers connected to corporate networks have become one of the most important productivity tools in the hands of chemists. A combinatorial chemistry, modeling and analysis environment that is built on top of this infrastructure enhances this. MSI's WebLab tools provide this environment, providing access to the tools traditionally reserved for experts, in an intuitive web-based environment and an efficient means of communication between the experimentalist and computational experts.
4:30 52 Parallel synthesis data handling via the Web.
David S. Nunn, Division of Chemical Sciences, Wyeth-Ayerst Research, 401 N. Middletown Road, Pearl River, NY 10965.
The increased complexity of combinatorial systhesis, relative to traditionsl medicinal chemistry, forces a greater reliance on computer tools for planning the experiment and tracking the data. Just as combinatorial chemistry achieves its efficiency through the reuse of chemicals, combinatorial computer tools must work from a description of the reagents involved and their description, rather than the products formed. We have implemented tools for interactive reagent selection, library design & visualization, systhesis tracking, and compound registration in parallel synthesis experiments. The web provides an attractive solution for delivering these tools to the chemists' desktop.



Marriott Copley Place, Salon C, 4th Floor

Recent Applications of 3D Databases
O. F. Güner, Organizer, Presiding
8:30 53 Exploiting receptor binding site using pharmacophore descriptors
C. J. Luttmann, Rhone-Poulenc Rorer S. A., 13, quai Jules Guesde B. P. 14, Vitry-sur-Seine Cedex, France, F-94403
Experience using 3D searching techniques based on pharmacophore queries has shown in the past to be very fruitful. We have analyzed the binding site of the metallo-enzyme Farnesyl Transferase in terms of all potential pharmacophores expressed and used this information to screen collections of compounds as well as virtual libraries for potential ligands and scaffolds, and also evaluate docking models. When, in addition, active compounds are known the task of finding potential novel leads is even facilitated by including SAR knowledge. Experimental design techniques can be used to reduce the number of potential ligands down to a manageable size in agreement with synthesis power while taking into account some constraints. In this talk we will present our experience using such techniques for library design under the active site constraints of Farnesyl Transferase.
9:00 54 Computational 3D searching and combinatorial chemistry combine to discover novel cyclooxygenase-2 inhibitor.
K. D. Stewart, L. Frey, E. Otis, V. Klinghofer, K. I. Hulkower, S. Loren, Abbott Laboratories, 100 Abbott Park Road, Abbott Park, IL 60064
Cyclooxygenase-2 (COX) inhibition is an area of current research for new anti-inflammatory therapy. We sought to discover inhibitors different from the current leads: Nimesulide and Celecoxib. A 3D searching algorithm, DOCK, was used to discover a phenothiazine ring system which can bind in the portion of the acitve site which is conserved for COX-1 and COX-2. Several analogs were tested that were equipotent against both COX-1 and -2 in the low micromolar range. From the proposed binding mode, extensions from the phenothiazine template were predicted to project into regions of difference between COX-1 and -2. A 40-member combinatorial library was designed to probe this possibility. While most of the compounds lost potency for both COX enzymes, several retained or gained affinity only for COX-2. The most interesting compound resulted from a facile 2-step synthesis, posessed no assymetric centers, had low micromolar potency against COX-2 and exhibited >50-fold enzyme selectivity.
9:30 55 New developments in pharmacophore model generation and in 3D searching.
Jon M. Sutter, Osman Güner, Hong Li, Marvin Waldman, Molecular Simulations Inc., 9685 Scranton Road, San Diego, CA 92121-3752
A pharmacophore is a representation of generalized molecular features that are believed to be responsible for a certain biological activity. A pharmacophore can be used to retrieve biologically active structures from a 3-Dimensional database or to predict the activity of new compounds. Catalyst/HYPO is a computer program that finds pharmacophores automatically based on how effectively the chemical features (e.g. Hydrogen bond donors/acceptors, hydrophobic groups, etc.) are able to predict the activity of a training set of compounds. It is conceivable that certain chemical features will have a greater contribution to the biological activity than others. To address this possibility, Catalyst/HYPO has been improved so that each chemical feature is allowed to have a different weight associated with it when the activity is computed. In this paper, we compare search results with and without using variable weights and discuss the implications.
10:00 56 Virtual high-throughput screening - an application of partial match 3D searching.
Trevor W. Heritage, Tripos, Inc., 1699 South Hanley Road, St. Louis, MO 63144
Recent rapid growth in the availability of 3D protein structures provides an exciting opportunity for the development of new drug candidates using structure-based design methodologies. 3D searching methods have primarily been used for ligand-based design and pharmacophoric searching. New types of 3D query features and constraints allow generation of receptor-based 3D queries, that can find molecules that are complementary to a receptor binding site. This presentation focuses on the combined use of 3D searching and docking as a strategy for screening large chemical databases for candidate hits, then ranking them in terms of receptor complementarity.
10:30 57 Parallel screening of HMG-CoA reductase inhibitors with Catalyst(TM).
David Jourand, Remy Hoffmann, 46 Allee d'Italie, 69007 Lyon, FRANCE
In the absence of knowledge of the receptor 3D structure, one approach in the drug design process combines pharamacophore generation based on experimental data activities from known inhibitors followed by screening of large 3D molecules databases to match potential hits. Because of the molecules' flexibility, one needs to build/generate for each molecule a conformational model that best scans its conformational space. CATALYST TM provides two conformational analysis methods : a slow but accurate method based on Poling used for hypothesis generation, and a fast but less accurate method using a systematic or poled search used to build databases. We investigate the influence of the conformational space coverage quality in databases on the discovery of new HMGCoA Reductase inhibitors. We propose a parallel and modified method to increase this quality, while building a 3D database in a tractable computing time. Finaly, we parallelize the 3D databases screening.



Marriott Copley Place, Salon C, 4th Floor

General Papers
C. Gragg, Organizer, Presiding
1:30 58 How electronic journals will change the way chemistry professors use the chemical literature.
Stephen A. Koch, Department of Chemistry, State University of New York at Stony Brook, Stony Brook, NY 11794-3400.
The advent of electronic-journals and the concurrent dramatic decease in the prices of hard drives, high speed duplex laser printers and high quality color printers has dramatically transformed my use of scientific journals. Previous to the existence of e-journals, I made Xerox copies of articles that were relevant to my research work. These articles were then arranged in folders in file cabinets. Now, after downloading and printing copies of e-journal articles, I store the pdf file of the e-journal article on my hard drive. Adobe Catalog (that is included in Adobe Acrobat Exchange) enables the creation of a searchable index of every word in all the articles in my personal database. The use of e-journals to create materials for teaching and research presentations will be discussed. And, the implications for university chemistry libraries regarding the transformation from the use of print journals to e-journals will be presented.
1:50 59 Publishing chemical information on the Web to support researcher buying decisions and chemical commerce.
Terry Graber, 18 Campus Blvd., Newtown Square, PA 19073
It is critical that researchers have quick and easy access to the most current chemical data and chemical ordering information. This case study will review the challenges faced by a leading provider of chemical products and services for research. It will show how data warehouses of chemical information were created to support researcher buying decisions and enable electronic commerce between customers and chemical vendors. The presentation will discuss the key information required by researchers, purchasing agents, safety officers, and storeroom managers. Lessons learned will be presented about how to organize web-enabled commerce to best serve researcher's needs. Also discussed will be the mechanisms and techniques used to streamline customer access to information thus shortening the supply chain between customers, suppliers and vendors.
2:10 60 The central chemical registry system at the Environmental Protection Agency.
Tommie G. Curtis, Lois Fritts, Marasco Newton Group, Ltd., 2801 Clarendon Boulevard, Arlington, VA 22201
The Chemical Registry System (CRS) is intended to be the central source of chemical information for the Environmental Protection Agency (EPA). The CRS will contain a unique identifier for each substance, a systematic name, a common name, synonyms, references to regulations which include the chemical, and, when available, a Chemical Abstracts Service Registry Number (CASRN). A user could query on a CASRN, systematic name, common name, synonym, or regulation. Search results could include the regulations in which a chemical is found, what chemicals are in a specific regulation, and detailed chemical identification information. The CRS will be the core of an information system that can link to other Agency databases. The system will be available to the public on the EPA web page.
2:30 61 Reinventing chemical information at the Environmental Protection Agency.
Lois E. Fritts, Tommie G. Curtis, EPA System Development Center, 200 North Glebe Road, Suite 300, Arlington, VA 22203
EPA has historically responded to Congress by creating independent Offices to enforce legislation. Each Office devised a schema for identifying chemicals: Water created parameter codes; Air, pollution codes; Solid Waste, hazardous waste codes. No process existed for aggregating data about a chemical across media. In response to programs such as Reinventing Environmental Information, the Agency established the Environmental Data Registry (EDR) as a data management tool. Data standards are recorded in the EDR and used when developing new or reengineered systems. The Chemical Data Standards Working Group, composed of chemists who represent Agency programs, is working with OIRM and state partners to establish a chemical data standard. The key data elements, valid domains of chemical identifiers, and business practices will be documented in the EDR and used to ensure Agency data sharing based on chemical identity.
2:50 62 Automatic classification of organic reactions - classification of reagents based on electronic and steric features of the three-dimensional molecular field.
Hiroko Satoh (1), Sachiko Itono (2), Kimito Funatsu (3), Keiko Takano (2), Tadashi Nakata (1);
(1) The Institute of Physical and Chemical Research (RIKEN), 2-1 Hirosawa, Wako, Saitama 351-0198, Japan.
(2) Ochanomizu University, 2-1 Otsuka, Bunkyo, Tokyo 112-8610, Japan.
(3) Toyohashi University of Technology, 1-1 Tempaku, Toyohashi, Aichi 441-8580, Japan.

A novel method named FRAU (Field-characterization for Reaction Analysis and Understanding), which numerically detects electronic and steric features of field around a molecule, has been developed for an automatic and systematic classification of organic reactions. Forty-two kinds of reagents, of which geometries and electronic structures were optimized by ab initio MO calculations were characterized by the FRAU. Using the features as discriminators, these reagents were classified by a Kohonen neural network. Good correlations were found among the features estimated by FRAU, similarities of substructures, and roles of the reagents in reactions.
3:10 63 AUncertainty analysis of reaction models based on the deterministic equivalent modeling method.
C. Wang, M. A. Tatang, and G. J. McRae, Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139
Uncertainty in modeling can originate from the structure and parameterization of the model, or from uncertain input parameters. Uncertainty associated with the chemical information, namely physical and kinetic properties, belongs to the category of parametric uncertainty. This type of uncertainty has been typically estimated using sensitivity analysis of sampling techniques such as the Monte Carlo method, which is not efficient for industrial-size reaction models. Instead, a computational efficient methodology termed deterministic equivalent modeling method (DEMM) can be employed to treat parametric uncertainty in reaction systems. Central to this meghodology is the application of polynomial chaos expansions and the probabilistic collocation approach to transform the stochastic model into a deterministic equivalent model. Examples to illustrate the application of DEMM in the uncertainty analysis of reaction models will also be presented.
3:30 64 ASLIMS: Sample, structure and spectral laboratory information management system with Web interface.
Antony Williams, Valeri Kulkov Advanced Chemistry Development, Inc., 133 Richmond Street West, Suite 605, Toronto, Ontario M5H 2L3, CANADA
At ACD we have designed and implemented a system to allow corporate- wide access to analytical information, focussing specifically on spectral information. In general, laboratory information management systems (LIMS) have failed to address the flexibility of interface and features required in analytical and R&D environments that include access to molecular structures and graphics intensive spectral displays. We have developed a web-based LIMS systems for managing spectral and associated molecular structure information. This user-friendly system links a unique sample identifier to sample information, a chemical structure or structures, associated spectra and final reports of analysis. This full-featured sample manager allows desktop access to sample information, structures and spectra, as well as access to a structure database.
3:50 65 Sharing a core chemical sciences collection among research libraries.
E. Jargo (Brown Univ., Providence, RI 02912), E.R. Lorbeer (Boston Univ., Boston, MA)
The Boston Library Consortium (BLC) is a union of 14 academic institutions in Massachusetts and Rhode Island. In 1997, a cooperative collections committee was formed to investigate a means to deal with the alarming increase in chemistry serial prices that is taking a toll on each library's ability to support current journal subscriptions. The goal of the committee is to ensure continued comprehensive coverage of chemical information throughout the consortium and provide opportunity for discussion of issues that significantly affect our chemistry collections. The resulting agreement calls for each BLC library to hold particular titles in order to maintain a "core" collection for the consortium. This paper outlines the methodology used to arrive at the agreement and demonstrates how cooperative collection agreements can work to preserve a collection in the chemical sciences. Finally, future problems associated with resource sharing, such as archiving, document delivery, and digital formats will be discussed.