annotateOl: Annotate the overlap between SVs

In jmonlong/sveval: SV evaluation

Annotate the overlap between SVs

Description

Overlaps that were prepared by prepateOl and then potentially filtered (by size, region of interest, call quality), are filtered to match at least a minimum overlap criteria. Different methods can be used to match variants, see below.

Usage

annotateOl(
  ol.gr,
  min.ol = 0.5,
  method = c("coverage", "reciprocal", "bipartite")
)

Arguments

ol.gr	overlaps prepared with prepareOl function.
method	the method to annotate the overlap. See details.
min.cov	the minimum overlap/coverage to be considered a match. Default is 0.5

Details

The 'method' is either 'coverage' (default) for the cumulative coverage (e.g. to deal with fragmented calls); or 'bipartite' for a 1-to-1 matching of variants in the calls and truth sets.

For some overlap methods (e.g. reciprocal coverage), a match is not always reciprocal. For example, a smaller deletion can be covered by a larger one while still not covering the large one enough. The means the small deletions is "overlapped" (here covered) but not the large one. To report this kind of overlap, two additional columns are added: 'queryOl' and 'subjectOl'. Hence, during evaluation it is not sufficient to cound SVs in the overlap object but rather the ones with queryOl==TRUE or subjectOl==TRUE. When using other overlap approaches (reciprocal overlap and/or bipartite clustering), both columns should always be TRUE.

Value

an updated and filtered version of the input ol.gr. The potential new columns include:

queryOl	should the query be counted as "overlapped"
subjectOl	should the subject be counted as "overlapped"
olScore	the overlap score (usually the value of the reciprocal overlap)

Author(s)

Jean Monlong

jmonlong/sveval documentation built on July 31, 2023, 7:50 p.m.