Just because you have a single transgene does not mean you cannot use standard tools. If you want the reads included mapping to it then add the cDNA sequence to the reference genome add make a new index. Or just use the standard index. It in this case does not really matter to align the potentially few reads coming from the transgene (if any, maybe the promoter). I like bowtie2 as it predictably and reliably uses a low memory footprint.