VarSeq 1.4.5 Release Notes
New Product Add-Ons
CNV calling designed for low and ultra-low read depth Whole Genome data is now available as an add-on licensed product to VarSeq!
Designed for calling large cytogentic events, this algorithm is able to detect chromosomal aneuploidy events with high confidence with ultra low read depth genomes (as few as one million aligned reads, or 0.02X coverage).
With higher read-depth Whole Genomes, the Binned Region Coverage can be adjusted to detect up to 10 events per sample, down to 100Kb sized events.
Both the WGS and Target CNV callers now also report a Karyotype for large events that span one or more cytoband region as well as a p-value to provide a confidence value on the CNV call.
CNV calling with the Target and WGS caller as well as computing the required coverage statstics can now be done from the new vscnv command line runner. See the VarSeq CNV Command Line Runner for more details.
CNV Caller on Target Regions has been updated to better handle large number of CNV events that occur in highly mutated tumor exomes. The detection of Duplicates was also finely tuned based on mutiple truth sets. Performance improvements came from better sample normalization techniques and fine tuning of the default model parameters.
Although some changes are to be expected for events very near the threshold of detection, high-quality events should continue to be consistently called. Nevertheless, as always, we recommend testing the changes to this algorithm against your own test set before moving it into production.
The LoH Caller has been updated with a new Triploid state called when the Hidden Markov Model is able to clearly detect a likely duplication purely based on the distribution of the Variant Allele Frequency. It is best practice to run the LoH caller before running the CNV caller on exomes, as the called LoH and Triploid regions are used to improve normalization during CNV calling.
You can now manage the samples used as potential reference samples for CNV calling in the Tools > Manage Reference Samples dialog. You can also add reference samples directly without calling CNVs on those samples add opting to add them as reference samples.
A number of customizations to the output of the Annotate Transcripts have been added. See the Annotate Transcripts algorithm for more details.
- An option to use the single-letter vs three-letter amino acid representation will customize the HGVS notation.
- A optional Preferred Transcript list can be provided to be used when selecting the Clinically Relevant transcript to display.
- On option to turn off the default filter to annotate only mature RNA.
You can now Lock the views of your project from the View menu, preventing them from being closed and moved. This can be used to ensure a consistent user experience when moving a Project Template into production.
- Features in a VCF file that had a ALT field of <CNV> and END value defined were being interpreted as GVCF span fields instead of being ignored on import.
- Certain field descriptions when exported to VCF had un-escaped newlines that prevented the VCF files from being used. This was resolved as well as other potential issues with special characters in data fields when exporting to VCF.
- The max operator in the Expression Editor would break when one data value was a missing value. Now the Max operator checks for missing values before taking the Max.
- When opening VarSeq through a remote SSH session from Mac to a Ubuntu server, the keyboard letter mapping was incorrect. Modifications of our Linux builds were done to correctly detect the keyboard mappings under this specific situation.
- The CNV , LoH and Coverage Statistics outputs tables can now display “All Samples” and not just the “Current Sample”. These and other options are available under the table toolbar Samples display options menu.
- The Ontology files for HPO, GO and OMIM were updated with data from May 2017. This is utilized in the PhoRank phenotype ranking and Match Genes Linked to Phenotypes features.
- There is now a Plot -> Plot Annotations… menu to it convenient to add a annotation source to a new or existing GenomeBrowse view as a new track.
- A number of functions were expanded to support working on the String Array type in the Expression Editor. For example, joindelim(regexplist( HGVScClinicallyRelevant , “(c\..*)”), ”; ”)could be used to create a HGVS list with the transcript names removed and joined by a semicolon.
- When editing a Filter expression for a GenomeBrowse Plot Item, the Positive and Negative examples for the current filter are now in tabs to allow for a more compact dialog.
- Added a new choiceText field type that can be used when Customizing Reports. This is especially useful for providing a dynamic list of choices based on the current context, such as picking a transcript to report for a variant.
- If vspipeline command was misconstructed, we now report back the command as part of the error message to improve the troubleshooting experience.