Skip to main content
. 2017 Jan 7;6(1):1–13. doi: 10.1093/gigascience/giw008

Table 2.

Predictive power of different morphometric methods to discriminate heteroblastic node

Heteroblasty Landmark EFD Both Max
1 49.1% 33.2% 47.9% Landmark
2 22.9% 19.5% 27.0% Both
3 12.3% 16.7% 15.3% EFD
4 13.3% 8.2% 12.7% Landmark
5 6.2% 12.2% 9.0% EFD
6 7.0% 9.0% 9.9% Both
7 15.9% 10.2% 10.7% Landmark
8 5.9% 13.6% 14.2% Both
9 10.9% 7.8% 12.0% Both
10 12.8% 11.1% 11.6% Landmark

For each heteroblastic node its percent correct prediction using the indicated morphometric features (landmarks, EFDs, or both) with linear discriminants is provided. “Max” indicates the set of morphometric features providing the maximum discrimination of heteroblastic node identity.