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 think it may be worthwhile to refactor combineBCR with some new argument like "cloneDefinition" where users can input different methods with the ellipsis argument ... to allow for parameters. While I have no issue with the current definition of clonotypes for BCRs, I can't help but feel it may be better for less technical users to incorporate some of the new methods in the future. There have been quite a few papers just from a quick googlesearch where different methods to cluster BCRs either independently or WITH scRNA data (this can be in a seperate function like inferBCRclonotypes or something) and those can be added.
API wise call.related.clones will be deprecated in favour of something like cloneDefinition="lvCompare" and threshold will be refactored into ... so not a huge change. This chaging process with be eased as well with the newly introduced lifecycle package in #417
Computational speed wise - parseBCR can still be re-done in rcpp I believe since the clonotype inference is done afterwards with lvCompare.
I think it may be worthwhile to refactor combineBCR with some new argument like "cloneDefinition" where users can input different methods with the ellipsis argument
...
to allow for parameters. While I have no issue with the current definition of clonotypes for BCRs, I can't help but feel it may be better for less technical users to incorporate some of the new methods in the future. There have been quite a few papers just from a quick googlesearch where different methods to cluster BCRs either independently or WITH scRNA data (this can be in a seperate function likeinferBCRclonotypes
or something) and those can be added.API wise
call.related.clones
will be deprecated in favour of something likecloneDefinition="lvCompare"
andthreshold
will be refactored into...
so not a huge change. This chaging process with be eased as well with the newly introducedlifecycle
package in #417Computational speed wise - parseBCR can still be re-done in rcpp I believe since the clonotype inference is done afterwards with lvCompare.
Thoughts? @ncborcherding
The text was updated successfully, but these errors were encountered: