A data-analysis pipeline for large-scale gene expression analysis

In this article we describe a method for characterization of large cDNA clone libraries based on oligonucleotide fingerprints (OFPs). The main advantage of this technique lies in that, without sequencing, each clone is tagged in an almost unique way, which has a couple of interesting applications, e.g. clustering of clones that belong to the same gene or gene family followed by sequencing of representative clones for each cluster. Moreover, small clusters are likely to represent rarely expressed genes, which are difficult to find by common approaches. We will demonstrate that in the EST projects carried out in our lab the global redundancy is very low compared to similar projects described in the literature, and simultaneously the number of unknown (novel) genes detected using this method is very high. In addition OFPs can be used directly for data base mining, since the sequences of the oligos matching a specific clone is known Recent results are presented, which underline the potential of our method in finding novel genes or genes homologous to known data. We will also address future applications in gene expression profiling, and give an outline of the various bioinformatics tools, which have been developed so far and which are used for automated data processing and analysis.