[Feature] Cluster significance testing using sc-SHC

[MicTott]

As @boyiguo1 suggested, we could potentially implement cluster significance testing using sc-HSC to determine which samples have artifacts.

After a quick glance, it seems like sc-HSC offers two possible solutions: 1) use their hierarchical clustering function scSHC to cluster out artifacts in PCA space, or 2) continue to use k-means/GMM and then use the their significance testing function testClusters to either accept or reject the cluster results.