Markus Riester
Markus Riester
setPriorVcf grew organically and cutoffs are ad hoc and might not work well for all combinations of germline and somatic databases and their corresponding versions. - [ ] Use population...
- [ ] All sample-specific annotation should go into FORMAT, not INFO. - [ ] Add all columns in the data.frame version to the VCF - [ ] Provide option...
We currently flag so many samples that this feature becomes rather useless. Need to become better at figuring out if a PureCN run was successful. - [x] Document flags in...
The current cutoff for homozygous deletions is large because we rarely see such huge deletions at MTAP/CDKN2A. Might be better to decrease the default and add a whitelist of known...
There are now a couple of third-party copy number visualization libraries available, e.g. http://bioconductor.org/packages/release/bioc/vignettes/CopyNumberPlots/inst/doc/CopyNumberPlots.html We should at some point switch to something better than our defaults.
We currently use flat priors for private variants. This can result in artificially high private germline rates in hyper-mutated samples or low rates in silent genomes, especially in high purity...
In high purity samples, PureCN currently isn't as good as it could be and frequently overestimates ploidy when copy number heterogenity is significant. - [ ] Find good test samples...
- [ ] Obtain test samples - [ ] Jointly infer M for phased SNPs in a segment - [ ] In predictSomatic VCF export, provide phasing in unbalanced segments.
In order to use PureCN's variant annotation in production settings, it would be better to flag filtered variants, not completely remove them. Low priority for now, since users can easily...
Hi, this might be a very niche problem, but I have datasets in which I'd like to rotate some samples by 90 degrees to align them better. I could not...