#221 - Pharma Informatics: Integration of Bioinformatics and Cheminformatics

221st National Meeting, San Diego, CA April 1-5, 2001 -- BTEC/CINF Abstracts

G. Grethe, Secretary General; G. Grethe, Program Chair


Pharma Informatics: Integration of Bioinformatics and Cheminformatics
Cosponsored with Division of Biochemical Technology
O. F. Güner, Organizer
1:00 32 PharmaInformatics. Perspectives on the integration and sharing of R&D data in a pharmaceutical/biotech environment.
Herschel J. R. Weintraub, Department of Scientific Computing, Genentech, Inc, 1 DNA Way, MS 56A, South San Francisco, CA 94080, Fax: 520-438-4295, weintraubh@att.net
Data integration requirements in pharmaceutical/biotech R&D environments present highly complex data management challenges. In Research, small molecule chemical structures, protein structures, genome sequences, annotation information, combinatorial library diversity and other properties, analytical and ADME/Tox data must all be collected and disseminated throughout the organization. In addition, imaging data, HTS biological assay data, and data from a myriad of LIMS Systems must be integrated. In Development, cell, tissue, animal, protocol, pharmacokinetic, QC, and other highly complex data is collected and must be shared with the organization without compromising the validated systems that contain them. Document management, especially relating to study protocols and regulatory submissions must be addressed. The use of clinical samples for R&D analysis presents another challenge regarding confidentiality of the data/tissue sources. An approach to undertaking such a massive data integration challenge is presented, without assumptions relating to legacy systems that may be in place in the organization. Examples will be given to illustrate key issues.
1:30 33 Informatics challenges in chemical data storage, retrieval and mining are being met with the development of new cheminformatics technologies and tools
Janet Cohen, Dave Diller, and Peter Gund, Pharmacopeia, Inc, 3000 East Park Blvd, Cranbury, NJ 08512, jcohen@pharmacop.com
The mass of data generated by current high-throughput discovery methods are overwhelming traditional cheminformatics systems. Fortunately, there are new back-end technologies for storing and representing chemical structures and chemical libraries, and new front-end technologies for accessing and analyzing the structures and data from the desktop. Now, data mining and analysis methods that previously were only available to the expert computational chemists are available on the experimentalists’ desktop. In this presentation we will explore some of the new technologies and tools available and present a case study on how some of these challenges are being met at Pharmacopeia.
2:00 34 Bridging cheminformatics and bioinformatics by using protein structures.
Ah Wing E. Chan, Molecular Design, Inpharmatica, 60 Charlotte Street, London W1T 2NU, United Kingdom, Fax: +44 20 7631 4844, e.chan@inpharmatica.co.uk, Roman A. Laskowski, Department of Crystallography, Birkbeck College London, and Janet M. Thornton, Structure and Modelling Group Biomolecular Structure and Modelling Group, University College London - SLIDES
A chemoinformatic database have been developed for generating drug design ideas of small molecules for given protein targets. The system relies on an automated method of identifying and delimiting known or probable binding sites and a knowledge based potential that describes favorable atom-atom interactions available within those sites. The potential is derived empirically from known protein and protein-complex structures and is based on 3D spatial distributions of atomic contact preferences between different atomic types. It contains no assumptions about energy functions and so reflects the way that atoms in proteins interact. The information on the target protein is further complemented by a comprehensive database of bioinformatics data holding sequence and structural alignments from all closely and distantly related proteins.
2:30 35 Mind the gap: Bridging the gulf between bioinformatics and cheminformatics.
William Langton, and Mike Higgins, Tripos, Inc, 1699 South Hanley Road, St. Louis, MO 63144, Fax: 314-647-9241, bill@tripos.com
In the field of life science research, significant resources have been invested to solve the challenges resulting from the huge quantities of data generated by genomics, proteomics, high throughput screening (HTS), and combinatorial chemistry. Bioinformatics groups have focused on genomic and proteomic data while cheminformatics groups have focused on HTS and chemistry data with little or no interaction. It has become obvious that researchers in their respective areas (genomics and proteomics, biology and chemistry) must make effective use of data from the other. Providing the tools for these researchers to accomplish this is the next challenge faced by informaticists. We will present ideas concerning the technology that can be used to bridge these islands of information.
3:00 36 Multidimensional exploration into biochemical pathways.
Johann Gasteiger, Dietrich Trümbach, and Wolf-Dietrich Ihlenfeldt, Department of Organic Chemistry, Computer-Chemie-Centrum, University of Erlangen-Nuremberg, Naegelsbachstrasse 25, 91052 Erlangen, Germany, Fax: +49-9131.8526566, gasteiger@chemie.uni-erlangen.de - SLIDES
Biochemical processes in living organisms are usually represented by complicated two-dimensional networks. However, the interrelationships in biochemical processes are multi-dimensional in nature. Access to this space can be provided by storing biochemical pathways in a reaction database. Searches can now be performed for names, full structures and substructures, reaction partners, enzymes and coenzymes, organisms, reaction centers, etc. By using a standard structure format other chemical databases and computer programs can be connected to this database. Furthermore, connection to bioinformation databases is provided. This database provides deeper insight into the mechanisms of biochemical pathways that also allow inferences on the metabolism of compounds. Furthermore, integration of this database into the information chain "genes - proteins - enzymes - regulation of pathways" can be envisaged.
3:30 37 Prediction of PGP transporter activity using calculated molecular properties.
Terry R. Stouch1, Olafur Gudmundson2, and Sean E. Ge2. (1) Macromolecular Structure and Biopharmaceutics, Bristol-Myers Squibb PRI, PO Box 5400, Princeton, NJ 08543-5400, Fax: 609-818-3545, terry.stouch@bms.com, (2) Bristol-Myers Squibb
P-glycoprotein (P-gp), a membrane transporter protein and a cause of multi drug resistance (MDR), is present in many tissues including brain and intestine. This protein transports many structurally diverse molecules out of cells and hence limits the efficacy of many drugs. Obviously, P-gp substrate activity is an undesirable ADME trait that is important to identify. Using a dataset of 98 compounds, we developed a QSAR based on physically meaningful descriptors that has an overall 72% prediction success rate, that for non-substrates was higher. Several forms of analysis verified the model and demonstrated a significant activity-related structure in the data space. Permutation test showed that the cross-validation result is significantly better than random classification and therefore, the developed model will have good predictive power. Descriptors of high leverage included polar surface area, molecular electrostatics, and molecular size and volume. Misclassified compounds were often large, conformationally flexible and identified as outliers. The descriptors used in the current model are quick to calculate and prediction can be done essentially instantaneously.
4:00 38 Informatics integration within the drug discovery pipeline at Arena Pharmaceuticals.
Gareth Jones, Arena Pharmaceuticals, Inc, 6166 Nancy Ridge Drive, San Diego, CA 92121, Fax: 8584537210, gjones@arenapharm.com
Using Constitutively Activated Receptor Technology (CART) Arena Pharmaceuticals is able to screen small molecules against orphan GPCRs. The drug discovery pipeline at Arena includes the identification of novel orphan GPCRs from the human genome; target validation; HTS and chemical lead expansion. Each of these areas requires considerable informatics support. Properly designed enterprise information systems can both meet this need and boost productivity by providing tools for decision support and real-time communication and feedback between the diverse groups of scientists (e.g. screening, molecular-biology, in-vivo pharmacology and chemistry). Arena has designed and constructed a database system with these criteria in mind: a web-based interface provides non-experts with easy access to all Arena’s scientific data. Different paths into the data allow users appropriate access: chemists have access to screening data for their compounds; molecular-biologists can view gene data and tissue distributions; the lab-scientist can quickly view the results of their experiment and managers can search for leads, generate reports and analyze data. Achieving this goal requires careful attention to users needs and the seamless integration of cheminformatics and bioinformatics.



Publish your own newspaper