Release notes

NewDefine compute resources per task run (API)

When creating a task via the API, you are now able to set instance type (top level) and maximum number of parallel instances for your execution without the need to create a new version of the app.

The following API calls have been modified to allow you to set or retrieve instance information for a specific task:

• Create a new draft task
• Modify a task
• Get details of a task

Support for compute resource settings per task run will soon be available on our visual interface as well.

ImprovementsFolders become a standard API feature

Folders are no longer in the Advance Access stage and are available as a standard feature on the API. Apart from not having to use the X-SBG-Advance-Access header for API calls related to folders, you are now also able to specify the destination folder within a project when uploading a file. Find out more in the Files, folders and metadata section of our API documentation.

Improvements Activate spot instances on project creation

When creating a new project through the visual interface, you are now able to set your preference for Spot instance usage in the Create a project dialog. This setting can later be changed from the project settings page, or overridden per task on the draft task page. Learn more about Spot Instances on our Platform.

ImprovementsData Cruncher – JupyterLab Beta

Data Cruncher is now using the latest release of the JupyterLab environment. All available preinstalled libraries have been updated and some new ones have been added.

We have also made more improvements in order to reduce analysis initialization time. It shouldn’t take more than four minutes to spin up your Data Cruncher analysis, regardless of the chosen compute resources.

Updates to the TCGA, TARGET, and CCLE datasets

As part of Seven Bridges’ ongoing partnership with the National Cancer Institute (NCI), authorized researchers can access valuable public datasets generated by the TCGA, TARGET, and CCLE initiatives through the Seven Bridges Platform, and Seven Bridges collaborates with the NCI Genomic Data Commons (GDC) on an ongoing basis to ensure alignment between the datasets available through the GDC and our Platform. In keeping with this, updated versions of the TCGA, TARGET, and CCLE datasets have been released on the Seven Bridges Platform. As of July 10, the legacy TCGA and CCLE datasets available through our Platform are fully aligned with those in the GDC Legacy Archive, and the TCGA GRCh38 and TARGET GRCh38 datasets are fully aligned with GDC Data Release 11.0. However, please note that this update is not currently available if you are using the Seven Bridges Platform on AWS EU as the cloud infrastructure provider.

As a result of these dataset improvements, some users who previously copied files from these datasets to their projects will no longer be able to access a subset of these files. In advance of the release of these new dataset versions, Seven Bridges contacted the owners of all affected projects. The users who were not contacted by Seven Bridges or by the owner of projects on which they collaborate can expect access to all TCGA, TARGET, and CCLE files in their current projects to continue without interruption.

ImprovementsSet null or empty values for app settings

When defining app settings prior to execution, you are now able to set null or empty values for the available inputs. This is possible using the two new buttons placed next to the inputs under the Define App Settings tab.

The Set null option is available both for simple inputs (such as strings or numeric values) and complex ones (such as arrays, records or maps).

Set empty is available as a button only for complex inputs (arrays, records and maps). For most of the simple app parameters, such as strings, empty value is set by simply removing the value from the input field and leaving the field blank.

We have also introduced support for displaying certain more complex input types, such as arrays of records, that are now also available through the visual interface.

NewVariant Browser (BETA)

Variant Browser is under active development with features being added successively. For any suggestions or bug reports, please contact our support team at support@sbgenomics.com.

Variant Browser is an application for genome analysis and interpretation that will allow researchers and clinicians to quickly annotate and accurately prioritize variants and genes involved in a disease. Variant Browser bridges the gap between sequencing/raw data management and clinical management, providing robust genome interpretation to expedite diagnosis and accelerate the understanding of the genetic basis of disease, drug response, and health.

During the BETA stage, you are able to interpret demo VCF files which are readily available inside the app. The available options include filtering with configurable filtering criteria, directly opening the Genome Browser at the position of the selected variant and displaying general statistics for the currently applied filters. In case you want to interpret your own VCF files please contact us at support@sbgenomics.com.

Future development will include an integrated flow for annotation and interpretation of VCF files using the Variant Browser, with annotated files provided by annotation apps on the Seven Bridges Platform (such as VEP annotation workflow or SnpEff), or uploaded by the users. Additionally, there will be more detailed and elaborate filtering options, along with more comprehensive display options for relevant analysis information.

Try the Variant Browser right now by accessing it in your project’s Interactive Analysis section and read more about it in the documentation.

Variant Browser resources

Several resources that will be used for creation of sqlite databases to be browsed further with Variant Browser are already available on the Seven Bridges Platform. Those are annotation workflows based on VEP (Variant Effect Predictor) and SnpEff. Workflows for annotation and conversion of a single VCF file to an sqlite database are: VEP annotation & DB conversion and SnpEff annotation & DB conversion, as well as Trios: SnpEff annotation & DB conversion or Trios: VEP annotation & DB conversion in case of Trios analysis. The workflows are available in the Variant Browser public project.

ReleaseGenome Browser official release

Genome Browser is no longer in the beta stage and has been officially released. The most important achievement in this release is guaranteed accuracy based on manual review of fields in the standard file compared to the most commonly used genome viewer.

Latest additions to Genome Browser also include a history of the last ten positions within the file, easier insertion of markers that takes less clicks than before and the ability to select a different reference file instead of the one that has been preselected based on the headers in the loaded BAM file.

We have also increased the maximum number of simultaneous BAM tracks, which is now up to twenty instead of the earlier maximum of only three. Higher number of tracks might affect the loading time, but will work properly as long as the browser memory limit is not exceeded.

To try out the new features, access the Genome Browser in your project’s Interactive Analysis section.

NewApp archiving

We have introduced the option to archive apps that you don’t intend to use in a project. Archiving will hide an app from the apps list, workflow editor and other relevant places across the Platform. However, this will not affect reproducibility, as you will be able to rerun existing tasks containing the app.

Archived apps can be displayed by changing the Status filter to Archived in the apps list within a project. To restore an app, click Restore next to the desired archived app.

Find out more about app archiving in our documentation.

Multi-Instance Whole Genome Sequencing GATK4.0 Workflow

We’ve published the Multi-Instance Whole Genome Sequencing BWA/GATK 4.0 workflow. This workflow keeps a similar price, with the improvement in total execution time of up to 3.0 times compared to a single-instance implementation. The Multi-Instance Whole Genome Sequencing workflow processes a 30x whole genome in as little time as 3 hours without any additional computational resource such as GPU or FPGA. The differences in precision, recall and f-score between multi-instance and single-instance workflows are lower than 0.001%, which is expected due to stochastic effects.

See the workflow on the Seven Bridges Platform.