You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have run into to some weird results when I had seqlets at the beginning of my sequences. Namely, when using additional_flanks = 3 with recursive seqlet calling, seqlets that with additional flanks intersected 0, were called as positive attribution seqlets, even though all their individual attributions were all negative. This behaviour disappeared when I set additional_flanks to 0.
I can try to provide some data to reproduce, but I think the issue could be in seqlets.py L 445 - 447:
Hey,
I have run into to some weird results when I had seqlets at the beginning of my sequences. Namely, when using additional_flanks = 3 with recursive seqlet calling, seqlets that with additional flanks intersected 0, were called as positive attribution seqlets, even though all their individual attributions were all negative. This behaviour disappeared when I set additional_flanks to 0.
I can try to provide some data to reproduce, but I think the issue could be in seqlets.py L 445 - 447:
Namely when start is 0, the index jumps to the last element. Or at least I don't see another place in the function that could explain this.
The text was updated successfully, but these errors were encountered: