Thank you for developing such a great tool!
Currently, when I perform sample-by-sample CNV calculations, I use a resolution of 0.0001. I wonder if I increase the resolution, which increases the number of subgroups, will it change the final infercnv result? Especially the file run.final.i...@expr.data, I will analyze this file later. These are my codes as follows:
infercnv_obj = infercnv::run(infercnv_obj, cutoff=0.1, # use 1 for smart-seq, 0.1 for 10x-genomics
out_dir="OC_CNV", # dir is auto-created for storing outputs
cluster_by_groups=T, # cluster
denoise=T,
num_threads = 30,
analysis_mode = "subclusters",
leiden_resolution = 0.0001,
output_format = "pdf",
#up_to_step = 15,
HMM=T )
Sincerely,
Lucas