automatcally cluster mitogenomes as well, to provide a backbone to map agains

[VascoElbrecht]

currently mitogenome sequences are provided, to make an alignment, and use that degenerated consensus as a backbone for mapping all reads against. It would be better to cluster the mito sequences first to prevent overrepresentation of certain taxa (this will be rarely the case and does not matter much, but does not hurt to implement this at some point).