Database and Analysis System for cDNA Clones Obtained from Full-Length Enriched cDNA Libraries

• Published in: In silico biology Vol. 2; no. 1; pp. 5 - 18
• Main Authors: Nishikawa, Tetsuo, Ota, Toshio, Kawai, Yuri, Ishii, Shizuko, Saito, Kaoru, Yamamoto, Jun-ichi, Wakamatsu, Ai, Ozawa, Masashi, Suzuki, Yutaka, Sugano, Sumio, Isogai, Takao
• Format: Journal Article
• Language: English
• Published: London, England SAGE Publications 2002
• Subjects: Amino Acid Sequence; Base Sequence; Cloning, Molecular; Databases, Nucleic Acid; DNA, Complementary; Expressed Sequence Tags; Gene Library; Image Processing, Computer-Assisted - methods; Molecular Sequence Data; Sequence Analysis, DNA - methods; Software; assembly; full-length cDNA; database; functional annotation
• ISSN: 1386-6338; 1434-3207
• DOI: 10.3233/ISB-00025

We have developed an efficient sequence-analysis system and a database system for clones
obtained from full-length enriched cDNA libraries made by using the oligo-capping
method.  We developed a semi-automatic analysis system for 5'- and 3'-end sequences.  It
pre-processes raw sequences (vector cut and accurate-sequence region extraction), clusters
the sequences, searches for similarities through public databases, annotates completeness
of clones and analyzes the ORFs in the sequences.  Newly developed or improved programs
are used in each step. A new program, ESTiMateFull is used to evaluate and to predict the
sequence-fullness based on comparisons with mRNA and EST sequences, respectively.  The
ATGpr program is used to predict sequence-fullness based on statistical information.  The
combination of full-length enriched cDNA clones and ATGpr fullness prediction resulted in
70% accuracy in the specificity and the sensitivity of the fullness predictions.  For the
ORFs predicted by the ATGpr, the signal peptides are predicted and a motif search is
performed by our new system.  We also developed a program that assembles our sequences
with dbEST sequences and developed a system to retrieve clones by the characteristics of
the ORFs.  As keywords, combination of various results of the analyses can be used for
retrieval.  And various results such as ORF features and database search results can be
shown on the same screen by multiple displays.  Full-length clones having interesting
functions can thus be retrieved efficiently by using this system.