search

etal / cnvkit

Copy number variant detection from targeted DNA sequencing
http://cnvkit.readthedocs.org
Other
540 stars 164 forks source link

Triangles for seg masked by pruning at 'ymin=-5' #645

Closed tetedange13 closed 3 years ago

tetedange13 commented 3 years ago

Should close #643, if this is well an unexpected behaviour (which I guess, but not 100% sure)

Have a nice day. Felix.

tetedange13 commented 3 years ago

Should close #643.

Default behaviour when using scatter --chromosome is to "prune" at 'ymin=-5' (see https://github.com/etal/cnvkit/issues/643#issuecomment-883936000) => This behaviour can cause probes to be masked (no concerns for probes) but for segments it can be problematic

With this PR, having masked seg raises a warning + they are signaled as triangles crossing Y-axis => X-pos of a triangle is determined by seg.start, so that number of seen triangles does not always match number of masked seg, due to 2 masked seg having same seg.start (which is the case for "amplicon.{cnr,cns}" - chr9)

See https://github.com/etal/cnvkit/issues/643#issuecomment-890461910 for an example obtained with "amplicon.{cnr,cns}" test files

Have a nice day. Felix.