doubt in mapping

[rathankar]

i did a mapping study using BOWTIE and TOPHAT for a sample read file:
1. The mapped statistics says:
number of genes in the mapping using bowtie: 47
number of repeat regions in the mapping using bowtie: 12
number of CDS regions in the mapping using bowtie: 42

while,
number of genes in the mapping using TOPHAT: 22
number of repeat regions in the mapping using TOPHAT: 6
number of CDS regions in the mapping using TOPHAT: 21

Why is this so. is there any relationship between these quantatities. i know that we use TOPHAT for mapping RNA-Seq data, but i did this for unmapped reads