Species-specific minimal intron sizes in humans, Arabidopsis thaliana, Drosophila melanogaster, and Caenorhabditis elegans are 92, 89, 61, and 48 bp, respectively [ 31], and Ex18 is much shorter.
SpliceMap was run with the same 6 kbp maximum intron size (the minimum intron size is not configurable).
For each gene, one of the two states (0 for minimal intron and 1 for size-expendable intron) was assumed and mapped onto the gene phylogenetic trees.
(iii) There were higher RDIs in those minimal introns with a size range of 88 124 nt than the narrower-ranged ones (50 86 nt), and the higher RDI suggested that minimal introns have a tendency to maintain an optimal size around 87 nt.
The stronger negative selection acting on the nucleotide content of protein coding sequences provides additional evidence to support the proposal that minimal introns in a size range of 50 86 nt are more likely to subject to purifying selection resisting to the overall deletion force acting in all introns.
Intron 2 had a minimal intron ancestor, but changed to size-expandable introns in Class II (92.5% supported) during the evolution.
While the minimal intron is universal and stringent in size, the fraction of minimal intron-containing genes appears stable within and across taxonomic scales [36].
In addition to having minimal intergenic spacing (median = 0.59 kb), intron size in Amphimedon is markedly smaller than other animals (see [ 5] and Table 2).
Both positive and negative selections can work on these indels; positive selection leads to the fixation of the minimal introns with optimal sizes and purifying selection tends to remove the minimal intron-containing genes (or the individuals in a population) that possessing over- or under-sized minimal introns.
When examining the dependence of size-effect on GC-effect, we observed that a majority of RDIs appeared against size-effect found in every minimal intron intervals we defined after removing the impact of GC-effect (Table 6).
