cross-sample panel

[GruA]

Hi,

I wanted to learn more about the output of makepanel subcommand of scan2.

Is a cross sample panel a list of potential germline/clonal mutations missed by GATK or more a list of consistent artifacts?
In other words, does it help to get rid of things that are real but not desired in the final list (get rid of not single cell specific changes) or does it help to lower number of false positives?

Also, is it used to filter snvs and indels or indels only (found in paper Fig.3)?