search

robertaboukhalil / ginkgo

Cloud-based single-cell copy-number variation analysis tool
qb.cshl.edu/ginkgo
BSD 2-Clause "Simplified" License
47 stars 28 forks source link

Lorenz-based cutoff metric? #23

Closed kunalkathuria closed 5 years ago

kunalkathuria commented 5 years ago

Hello,

I was wondering if there is a built-in metric (or e.g. a way to use the normalized bin counts post-analysis) to identify "bad cells" (similar to having a BIC score for each cell) ?

Thanks,

Kunal

robertaboukhalil commented 5 years ago

A good starting point would be the SegStats file, e.g. here. I don't think we currently output GC stats + MAD values in the GUI, but if you run it through the CLI, you could probably infer those values and use them for quality check. Hope this helps!

kunalkathuria commented 5 years ago

Thank you! I may also be looking at Lorenz curve deviations for each cell.