Table 1.

Distance correction provides the best results on real CRISPR-Cas9 data

clone 1clone 2clone 3clone 4clone 5clone 6clone 7clone 8clone 9clone 10clone 11clone 12clone 13clone 14clone 15clone 16clone 17clone 18clone 19clone 20
HD452918049174101446605245518573827274160673811283682214837128735582100431
Corrected HD446117939094049423591239213413673262149571911253272065883124632912013390
WHD449618079154045440587249116853907257161469610123212177821134132502191410
Corrected WHD444317769013950423571239013993696260151770110453061990791111031202073390

[i] On real data from a mouse model of lung adenocarcinoma (Yang et al. 2022), our distance correction method improves performance as measured by the (Camin-Sokal) parsimony score of the reconstructed trees. The best method for each clone is shown in bold. Specifically, on 17 out of the 20 clonal populations the best results are obtained by using distance correction.