Closed ceheitsch closed 4 years ago
@ceheitsch You didn't provide the link for the clique2.dot file. I am currently unable to test out and resolve the issue with the special case you pointed out above.
Sorry about that! It seems that I loaded the nopct file twice. Clique2.dot now included.
clique2.dot.txt
@ceheitsch I'm not entirely sure I'm understanding the split problem well. Does this fix it?
Seems to be fixed now.
The T99.dot prediction has 12 TP, 14 FP, and 10 FP when compared to the native AE014184_h.nopct structure. The T99.dot and clique2.dot predictions are the same. The two pairwise comparisons are correct, but the counts in the triple comparison have the 14FP base pairs split differently between T99.dot and clique2.dot structures.
AE014184_h.nopct.txt T99.dot.txt AE014184_h.nopct.txt