Abstract

Expressed Sequence Tags (ESTs) are tiny pieces of DNA sequences that are generated by sequencing from either the 5 ' or 3 ' end of an expressed gene. They provide a highly cost-effective method of accessing and identifying expressed genes. The large-scale EST sequence data, generated in the Concordia fungal genomics project, raise the following two critical questions that need to be answered by their bioinformaticians. First, from the biologists' point of view, what is the most useful information encoded in these sequences, and how can that information be obtained? Second, what is the most efficient way for the scientists to access this kind of information? In this thesis, we provide our answer to the above questions by building an EST annotation system that can fulfill the tasks for EST data storage, function assignment and visualization. Annotation is the processes of adding biological information to EST sequences. The contents of an EST sequence annotation system could vary from one system to another; however, based on an actual genomics research project, our EST annotation system not only presents most of the common features in genomics annotation systems, but also includes a number of specific functions that are very useful for EST projects, such as sequence full-length prediction, sequence-based and unigene-based data integration, and protein signal peptide prediction. The methods used and described in this thesis could be a very useful protocol for most EST sequence annotation systems