A method to detect gene structure and alternative splice sites by agreeing ESTs to a genomic sequence

In this paper we propose a new approach to the problem of predicting constitutive and alternative splicing sites by defining it as an optimization problem (MEFC). Then, we develop an algorithm to detect splicing sites based on the idea of using a combined analysis of a set of ESTs alignments to a genomic sequence instead of considering single EST alignments. In this way we require that all ESTs alignments must agree, i.e. are compatible to a plausible exon-intron structure of the genomic sequence. Indeed, we show that a progressive and independent alignment of ESTs may produce unsupported splicing forms. Our method has been implemented and experimental results show that it predicts alternative splicings with high accuracy and in a small amount of time. More precisely, compared to published splicing data, the method confirms validated data while in many cases it provides novel splicing forms supported by several ESTs alignments. © Springer-Verlag Berlin Heidelberg 2003.