■ produces a classification based on shared derived similarities and reflects phyletic relationships |
■ produces a classification based on overall similarity and may not reflect phyletic relationship |
■ uses one algorithm for the analysis of all types of cancers |
■ may require a specific algorithm for each cancer type |
■ discriminates between ancestral and derived states; uses only derived character states (apomorphies) |
■ does not discriminate between ancestral and derived character states; uses both |
■ resolves issues of parallelism (multiple independent origins) by parsimony or maximum likelihood |
■ does not resolve issues of parallelism |
■ offers predictivity |
■ does not offer predictivity. |