The approaches for discriminating between non-coding and coding sequences resemble the methods for gene discovery in genomic sequences. They are based on either similarity to known coding sequences or the statistics of codon frequencies specific to each organism. Similarity to known coding sequences is detected by using tools for homology search, typically BLASTX [6]. It should be noted that the existence of even a short functional segment of peptides (motifs) supports the
identification of the transcript as coding. For novel lncRNAs, however, we cannot always expect conservation between species. The simplest way that does not rely on phylogenetic conservation is to check whether there is a long open reading frame (ORF) where no stop codon appears. However, it is necessary to evaluate any candidate segments more carefully because there are short ORFs that encode functional peptides [7].