This can be broken down into two scenarios:

a. When utilizing the --intron excluded parameter, a read is only counted if it aligns to the exon region of a gene, specifically when more than 50% of the read's length aligns with this region. Reads containing sub-regions or those aligning with intergenic regions are not counted.

b. When the --intron included parameter is used (this is the default setting), a read is counted if it aligns either with the exon or the intron region of a gene, as long as over 50% of the read's length aligns with these regions. Reads aligning with intergenic regions are not counted under this parameter.