mutation_burden format for stats2json and stats2lims

[bryce-turner]

For mutation burden we can only get sample level mutation burden, but we report it at the library level for consistency with other metrics tools. As a result we currently are using the first library in the list when jetstream parses the datafiles. We should duplicate the mutation burden for each library:

For example:

## METRICS CLASS
MutationsCount  Coverage        MutationBurdenPerMillionBase    SAMPLE  LIBRARY READ_GROUP
68     34046067        1.99729        EXAMPLE_1_BM_T1      K12345
68     34046067        1.99729        EXAMPLE_1_BM_T1      K12346