1. Home
  2. VarSeq
  3. VarSeq Release Notes
  4. VarSeq 2.0.2 Release Notes

VarSeq 2.0.2 Release Notes

New Features

  • In VSClinical, variants can be added to an evaluation from the project as well as from the user provided HGVS description of any variant. This allows adding variants sequenced outside of the test, correcting variant representation or just exploring a novel variant from within VSClinical. From VSClinical with an evaluation open, click the Variant List and then “Add Variants” to open the new dialog.
  • VSClinical now allows you to specify the thresholds used by the three population frequency criteria BA1, BS2, and PM2. The ACMG Classifier algorithm also allows these parameters to be set, in which case the will be used in subsequent VSClinical windows. They can be configured in the options dialog of the VSClinical tab and the relevant thresholds are displayed at the top of the Population tab for the selected gene disorder inheritance model.
  • New templates for gene panels and trio analysis have been added that use the ACMG Auto Classifier as the primary mechanism to prioritize variants for filtering.
  • In an ACMG Evaluation, you can now dismiss a variant that failed QC or otherwise should not be saved to the evaluation knowledgebase.
  • You can now open a finalized evaluation in VSClinical in read-only mode.
  • For the population frequency criteria, VSClinical now evaluates the minor allele, which is not necessarily the alternate allele. The max sub-population frequency reported is also now presented as a MAF.
  • During the import process, the types of custom VCF fields can now be adjusted during import when selecting the Advanced Options checkbox. This allows, for example, for custom fields to be converted to Categorical to aid filtering.
  • For VarSeq projects, you can now delete algorithms and sources in the project that other algorithms depend on. All outputs and downstream algorithms that will be deleted are listed in the confirm delete dialog.
  • Novel Splice Site detection has been added to the Annotate Transcripts algorithm when selected by the user as an optional output. Variants that introduce a new splice site or alter the motif of a nearby splice site to flip an algorithm to detecting it will be reported.

Bugs Fixed

  • The CNV on Target Regions algorithm was updated to handle a couple very specific input data issues that caused the output CNV events to not be the correct size. In one instance, large events where potentially getting truncated to where there was a dramatic change in Z-scores within the event. In another case where large number of targets where in the considered windows, legitimate events were extended erroneously beyond where the algorithm detected the signal.
  • Custom display preferences of table fields in project templates saved in 2.0.1 were reset when the project template was used. This issue has been resolved and all table display preferences are now preserved.
  • In VSClinical, the mutation hot spot criteria question was being displayed for all variant types instead of just for missense variants. It is now hidden for non-missense variants unless the option to show hidden questions is enabled.
  • When working with trios in VSClinical, the de Novo state was not being considered for projects where the parent genotype was missing. Now the PM6 code (unconfirmed de Novo) is used in this situation.
  • A recent regression was fixed that caused an error when adding a new Compound Het algorithm to a project and skipping the optional allele frequency input.
  • The current sample name in VSReports will now update immediately if you use the new File > Set Sample Fields features to change the sample name.
  • With certain user data, exporting the coverage table from the project as an annotation would auto-detect an incorrect genome assembly. Now the genome assembly set for the exported annotation always matches the opened project.
  • In GenomeBrowse, filters applied to plotted values are now also being applied to the Feature List table.
  • Running the Sample PhoRank algorithm will show the results in the Variant table. In a recent regression, these outputs were hidden when adding the algorithm to a project.
  • Algorithm process flow is more systematically controlled by not activating dependent algorithms until the current algorithm is finished running. This means the prompt for an algorithm input from an unfinished algorithm will not be displayed.

Polishes

  • The system default inheritance model for genes has been updated to reflect consensus data from OMIM Phenotypes 2018-07.
  • For the ACMG Auto Classifier and the VSClinical recommendation engine, UTR variants are included in the variant types used to recommend BP7. This improves the use of the auto classifier as a variant filtering utility.
  • The Golden Helix pathogenicity calculator has more intuitive starting states for criteria such as BA1 that were previously swinging the score unexpectantly when changed from Unanswered to No.
  • On Mac, the ability to have Assessment Catalogs connect to MySQL has been simplified by bundling the appropriate MySQL driver. This removes the requirement of installing a third-party driver in the correct location.
  • Also on Mac, opening VSClinical or a Web Browser tab added multiple VarSeq icons to the system dock, while only one of those was useful for switching the current application. This has been resolved.
  • When using the VarSeq configuration dialog, a warning will prompt when clicking OK while the AppData folder was changed but the required Move action was not yet taken.
  • The output from the ACMG Classifier now outputs the user-settable Gene Inheritance for the current gene as well as the transcript strand direction (+ or -).
  • The ACMG Classifier capability has been expanded to more easily handle GRCH38 and explicitly not allow GRCh37 hg19 human genome sequences.
  • In VSClinical start screen for the ACMG guidelines, you can delete an evaluation with the new trash icon without opening it.
  • VSClinical now has a spinning indicator as it loads a variant and does so in a single bulk operation compared to the previous behavior of many small parts of the screen changing individually and not in a synchronized manner.
  • When importing VCFs that do not specify the AD (Allelic Depths) field but do provide other fields used to add an auto-computed Variant Allele Frequency (VAF) field, an auto-computed AD field is filled into the project and can be used by VSClinical to display the reads attributed to the reference and alternate alleles for a variant.
  • The annotate transcript algorithm HGVS for frameshifts has been updated to include the optional length to the new stop codon such as NP_001035091.1:p.Arg83Glnfs*6 instead of just NP_001035091.1:p.Arg83Glnfs
Updated on January 21, 2021

Was this article helpful?

Related Articles

Leave a Comment