I am using TopHat to map RNA-seq reads. My study species is a non-model organism, so I created a custom genome build in order to do this. The first time I tried to map, I got an error message (reference sequence has more than 2^32-1 characters), so I split the genome in two, created two builds and mapped the reads to each build separately. I now have two accepted_hits BAM files that I want to join before performing further analyses. I tried using SAMTools > Merge BAM Files (Galaxy GVL-QLD mirror) but the job runs for a week without being completed. I think it might be related to this issue https://github.com/samtools/samtools/issues/260 , as the reference genome contains over 1 million contigs, but I'm not sure. Can anyone suggest a work around? I would prefer to continue using the Galaxy GUI rather than the command line version of SAMtools if possible.