Yes indeed. And new tools are being written - see the Adam project for an interesting example: https://github.com/bigdatagenomics/adam and the associated variant caller Avocado: https://github.com/bigdatagenomics/avocado. Others are also trying to get the old tools working on Hadoop, for instance Halvade: https://github.com/ddcap/halvade/wiki/Halvade-Manual, Hadoop-BAM https://github.com/HadoopGenomics/Hadoop-BAM, SeqPig: http://seqpig.sourceforge.net/, and the guys at BioBankCloud: https://github.com/biobankcloud. It's going to take quite a while for this stuff to get fleshed out, and for researchers to adopt it. But the sheer weight of data is going to force things in the Hadoop direction eventually. It is inevitable.