Parallelize correlation clustering

[sean-la]

To decrease runtime and take advantage of multicore processors, I suggest parallezing correlation clustering so that the clusterings of each genotype (e.g. MLST, CNV, SNP) all be performed at the same time.

[sean-la]

This could also be extended to include the entire pipeline, i.e. parallelize the process of genotyping, distance matrix calculation, and correlation clustering over multiple CPUs. Only the consensus clustering phase cannot be parallelized.