Effective clustering for single cell sequencing cancer data