It is likely that these delay kinetics regulating transcription, nuclear export and/or translation are different for differently sized genes (Notch1 has 33 introns, 34 exons and a transcript length of 9488 bp, whereas Dll1 has 10 introns, 11 exons and a transcript length of 3581 bp).
For example, the breast cancer cell line cohort included two splice site mutants that escaped detection by PAC because each had a second transcript length of major intensity that resulted from cryptic splicing (BRCA1 c.5396+1G>A in MDA-MB-436 [14] and p16 c.150+2T>C in MDA-MB-436 (Naget et al., submitted for publication).
Novel transcripts must (1) not overlap with any known RefSeq or UCSC gene exon, (2) have a total transcript length greater than 3 kb, (3) expression in either alpha or beta cell experiment > 1 RPKM.
We only considered genes with a sufficiently high expression level in at least one of the 8 samples (>2 Reads per Kilobase per Million of Reads, RPKM), which change their expression by at least 2-fold, and have an mRNA transcript length of at least 100 bases.
In this data set, some genes, such as MR1, have long transcript length with many exons, and an isoform with a small portion of exons is expressed.
First, we provide a carefully considered interpretation of the functional potential of the model; e.g. whether the transcript has a full-length CDS, is subject to NMD or non-stop decay (NSD) (25), or has a retained intron that would disrupt the CDS.
Having demonstrated that transcript length and AA AG ratio regulate protein levels in an Elongator-dependent manner, we next investigated the identity of specific Elongator targets.
Other studies have also considered transcript length in the context of translational efficiency and gene function [ 8, 45].
However, only 1 gene in chimpanzee and only 46 genes in RM have significantly different transcript length distributions or numbers of exons.
Also, I have extensively analyized CAGE data, which has the advantage that it quantifies directly the expression of a TSS (i.e. without having the complication of having different transcript lengths) and the distribution of expression levels of TSSs NEVER look bimodal.
