Extending the Domain of Comparative Genomics

Comparative methods have 2 obvious limitations at present that easily can be identified: Their dependence on an alignment and the assumption of a common structure shared by the compared genomes. These limitations have created a Scylla and Carybdis of comparative genomics: If the sequences are too close, the genomes are easy to align and the gene structure probably hasn't changed, but there are too few evolutionary events for the comparative principle to be exploited maximally. If the genomes are too distant, the alignment becomes unreliable and the gene structure could have changed, but there will be plenty of evolution that could distinguish hypotheses about gene structure. We will here propose a model that could alleviate both these issues in the case of gene finding and improve comparison of more distant genomes.