A system is described for cruising the sea of information in the world of human genome. The system named GenomeCruiser combines at any start-up personal data with the data from GDB. Therefore, it is a convenient laboratory notebook for the user to simultaneously access his/her experimental data and the most up-to-date public data. GenomeCruiser is also equipped with a friendly and powerful interactive user interface.

View full abstract

A solution system for the bottlenecks of a large-scale DNA sequencing project is now under development in the Kazusa DNA Research Institute. The system provides an integrated sequencing project management capabilities. Its functions are ranged from monitoring of running equipments, viewing of every data flow of DNA clones, contigs and raw data from automatic sequencers, commands for setting instructions to specified clones, a multiple alignment editor, to sequencing data quality control functions.

View full abstract

An animal genome database for network users was developed. Linkage map, cytogenetic map, reference, PCR primer information and homology between species were included in the database. These data can be used by electric mail system and World Wide Web client.

View full abstract

The Human Genome Project activities have been resulting in data accumlation on various databases. Each of them consists of a large amount and variety of data. By using these databases cooperatively, researchers can get information more widely and deeply.
Relational database management systems have been used for one of methods for integration of these databases. These database systems, however, require that data be represented as groups of two-dimensional tables, with the relationships among these tables defined by special columns and separate tables. Because of this restriction, users are forced to contort data into the required form.
By contrast, object-oriented database management systems let users model information in structures that are natural to the data and the application. These systems can manage complex data powerfully and integrate several databases easily.
We have been developed protein variant database collaborated with JIPID, Japan node of PIR-International, and also developed protein database query system. In addition, several novel databases are developed by JIPID in collaboration with National Biomedical Research Foundation (NBRF). All these databases as well as the sequence databases will have interfaced data items, so data that exist in one database may be transferred to another database. These databases are collectively called the database web. In order to utilize these database more effectively, we have attempted to develop an integrated protein database system using an object-oriented database management system. In this system, we treat PIR protein sequence database, variant database, enzyme database, and protein secondary structure database. Last three databases are developed by JIPID.

View full abstract

Computational methods are indispensable tools for researchers who determine a new nucleic acid or protein sequence. Such methods are used to obtain informaion to help identify and annotate the new sequence. Database searching is currrently the most popular way to get this information. Using homology- search programs such as FASTA or BLAST, one can search a database for sequences similar to a given query sequence. In addition, the alignment of multiple nucleotide or amino acid sequences is important for drawing phylogenetic trees and for predicting the structure of proteins. The multiple sequence alignment of proteins is also used to discover motifs and biologically important patterns.
We have developed a retrieval system for nucleic acid and/or protein sequences. The system is an Xtpanel application that runs on Sun Workstations with the X-Window system. Xtpanel, which is written by Steve Cole and Dave Nichols of Stanford University, allows one to build interactive X-Window system interfaces without having to do conventional X programming.
It supports FASTA and BLAST for homology searches, bfind and bget for database entry retrievals, and CL USTAL V for multiple sequence alignments.

View full abstract

Smith and Waterman dynamic programming algorithm is too computer intensive to be used for database search. Most of the common database search programs us approximate scores to reduce search time in stead of Smith and Waterman score. They do not consider all possible alignments and may miss some sequence similarities. We present an algorithm for database search which uses score table of k-tuples. The method calculates the upper limit of Smith and Waterman score for each database sequence in nearly interactive time. Calculated each upper limit of Smith and Waterman score prevents us from missing high sequence similarities.

View full abstract

We have been developing a protein structure database system named PROTEIX. In this report. we describe recent improvements on PROTEIX.

View full abstract

GEISHA (Genomic Environment for Interaction Simulation and Hypothesis Approximation) system verifies and revises the rules of pattern formation in Dorosophila embryogenesis. Thesystem consists of three parts: rule-based simulator, evaluator, and user interface. The simulator tests all the possible rule patterns, and the evaluator qualitatively evaluates results of the simulator. The user interface is named SIMFLY and provides a real-time graphic viewer and a rule-controller. We can do most of these operations using GUI.

View full abstract

We developed a knowledge base for searching and browsing metabolic pathways. Using this system, we can retrieve various pathways related to genetic diseases, chemical compounds, and enzymes.

View full abstract

NCBI maintains a database combining biological sequences and associated Medline citations. This information may be accessed over the Internet via an easy-to-use browser, Entrez. With Entrez one may rapidly discover what is known about a biological molecule, exploring “links” between sequences and citations, the homologous “neighbors” of a sequence, and “neighbor” citations, which discuss the structure and function of related molecules. A recent addition to Entrez is information describing 3-dimensional structure, allowing one to examine directly the structure of a biological molecule or its homologs.
The 3-dimensional structure database used by Entrez is called MMDB, for Molecular Modeling Data Base. It consists of messages in the ASN.1 language, which are translated automatically to in- memory data structures using C routines available in the NCBI toolkit. The MMDB data specification gives a complete description of the chemical structure of a macromolecular assembly, with unambiguous linkage to atomic coordinates. This data organization is intended to facilitate the computations required for homology modeling, which is based on alignment of the residue sequence, or, more generally, on comparison of the chemical graphs of different molecules, including their non-polymer components.
NCBI supports an active research program in comparative molecular analysis techniques which make use of 3-dimensional structure information. Computational methods for sequencestructure “threading” provide a means to generate model structures for proteins which are very distantly related to a protein of known structure. Methods for rapid structure-structure comparison provide a means to detect architectural similarities and/or distant evolutionary relationships among proteins of known structure. These methods currently make use of the NCBI “PKB” research database. In the near future, however, they will operate as client software accessing MMDB, and in this way provide examples of how applications developed elsewhere may make use of this resource.

View full abstract