GRanges / IRanges maxgap behaviour change raises no warning
1
0
Entering edit mode
@daniel-cameron-6227
Last seen 8.0 years ago
Australia

findOverlaps documentation mentions that G/IRanges is moving a Nested Containment List objects implementation with slightly changed behaviour. The documentation mentions that special meaning of maxgap in conjunction with start/end/within is not respected in the new algorithm. What is alarming to me is that there is no error or warning given for the new behavior. Shouldn't an error (or at least a warning) be raised given the nature of the change and the high likelihood of an unnoticed change analysis results?

Edit: additionally, different results are also returned for type="equal", which is not mentioned in the documentation of the change

 

 

genomicranges granges iranges • 876 views
ADD COMMENT
1
Entering edit mode
@herve-pages-1542
Last seen 17 hours ago
Seattle, WA, United States

Thanks Daniel for pointing this out. I agree that this kind of subtle change in semantic deserves a warning. I just added the warning to IRanges and GenomicRanges in release and devel, and also added the type="equal" case that was missing to the documentation of the changes between the old and new implementations (in ?NCList).

Cheers,

H.

ADD COMMENT
0
Entering edit mode

FWIW, the new implementation of findOverlaps/countOverlaps now also supports maxgap's special meaning when type is set to "start" or "end". So no more warning in that case. This change is in the release and devel versions of IRanges and GenomicRanges (IRanges 2.2.4 & 2.3.11, and GenomicRanges 1.20.5 & 1.21.15).

H.

ADD REPLY

Login before adding your answer.

Traffic: 1074 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6