Fix pseudobulk plotting legend (#182)
Fix Numbat$plot_exp_roll (#169)
Fix CNV states reporting when segs_loh
is provided (#183)
Fix n_states
reporting (#178)
Improve error handling in pileup_and_phase
(#179)
Integration with hahmmr
Better input checking for pileup_and_phase
Fix compatibility with igraph v2.0+ and tidygraph v1.3+ (#150)
Fix multiallelic CNV state probability reporting (#146)
Fix plotting issue #135
Fix CRAN check compilation issues
segs_loh
/call_segs_loh
enabled.Allows users to supply existing CNV profiles (e.g. from bulk WGS/WES analysis) via segs_consensus_fix
parameter
Adding call_clonal_loh
option to call clonal LOH events within run_numbat
Fixing bug #81
Fixing oversegmentation issue in find_common_diploid
caused by annot_segs
Introduce n_cut
parameter to specify the number of clones to define from the phylogeny
Allows users to redefine subclones from the phylogeny via nb$cutree
Numbat now works for F1 hybrid mice! Check out the new tutorial under Articles
.
Offers stacked clone bars in plot_phylo_heatmap
Externalize phylogeny module as separate package (scistreer
)
Prepare for new CRAN version
Better CNV state legends for plot_bulks
Improving error handling and removing python dependency (argparse
) in pileup_and_phase.R
Allows plotting of mutliple annotations in plot_phylo_heatmap
(thanks to @whtns)
Adding diagnostic messages
Fail gracefully when no CNV remains after retest_bulks
Passing gamma
parameter to retest_bulks
Conform to CRAN guidelines
Removed ATC2 examples from package data - users can download from lab server link instead
New option to specify genome version (genome = 'hg38' or 'hg19'
). Support plotting of centromeres and gap regions for hg19.
Removed genetic maps from package data and they are no longer provided as input to run_numbat
. Annotation of genetic distance is performed in pileup_and_phase.R
script instead, using the genetic map included in Eagle2.
Speed up of NNI using RcppParallel (#34). 10x faster and much more memory efficient (memory requirement is constant with respect to the number of threads).
Speed up of expression single-cell testing using roptim. Approximately 2x speedup.
New LLR metric for CNV filtering that is not inflated (default: 5).
Only keep heterozygous SNPs in alelle dataframe to reduce memory usage