CopyKit 0.1.2
New features
- New cell smoothing method with the function
knnSmooth(). Uses k-nearest neighbors to smooth cells profiles and re-segments the datasets obtaining cleaner copy number profiles, with reduced overdispersion and improving downstream analysis. (Thanks to Runmin Wei for the helpful discussion.) - scquantum method is available for
calcInteger()and is now a CopyKit import (scquantum is a single cell ploidy estimation tool developed by Alexander Davis) calcInteger()now accepts option methdo = 'metadata'. To use this option the user can add custom values of ploidy to every cell in the colData column 'ploidy' and runcalcInteger(ck, method = 'metadata')to obtain the integer matrix on the CopyKit object.- runVst allows the selection of the assay for the transformation
Changes
-
plotHeatmap()order_cells argument now defaults to NULL. NULL option respects the order of the CopyKit object. order_cells argument can be set to 'consensus_tree' and 'hclust'. -
Method 'scquantum' from
calcInteger()adds 3 elements to the colData.
- ploidy: contanining the inferred ploidy call for each cell
- confidence_ratio: ratio from scquantum inferred ploidy to scquantum theoretical ploidy
- ploidy_score: Score derived from the confidence ratio. Values closer to 0 indicate a better fit of the ploidy call
- Significance thresholds for CBS alpha segmentation and Merge levels were reduced to increase sensitivity to focal amplifications.
Removed
- option 'phylogeny' from function argument `plotHeatmap()' 'order_cells' has been removed.