I have been reconstructing novel coding or non-codng transcripts using either Cufflinks or Scripture from 2 years. They are very efficient tools to do de novo transcriptome reconstruction. However, sometimes both fail to do the job though there is more than enough RNA-Seq read coverage in an unannotated genomic region. I even noticed computationally predicted splice donor and acceptor sites overlapping these RNA-Seq reads. These features confirms that there is an expressed transcript but just missed by either cufflinks or scripture.
Now, is there anyway, I can manually design my transcript exons and introns using RNA-Seq mapped reads of a given genomic region ? I would really appreciate your suggestions on this.
Thanx in advance.