v5.13

• Replace GRIDSS somatic filtering with GRIPSS 1.3
• Upgrade Bachelor to 1.10
• Upgrade Purple to 2.45
• Upgrade Linx to 1.10
• Upgrade Amber to 3.4
• Make pipeline retry on all Quota failures
• Add public profile to run pipeline externally to HMF

https://hartwigmedical.atlassian.net/issues/?filter=10370&jql=fixVersion%20%3D%205.13&atlOrigin=eyJpIjoiNzAyMTk4MDI1OTM2NGM2ZjgyOTJiNDgzZWI5ZjMyZTEiLCJwIjoiaiJ9

v5.12

Features:

• Use SAGE (v2.2) as somatic SNV, MNV and small indel caller.
• Upgrade GRIDSS from 2.8.3 to 2.9.3
• Remove bcl conversion (now in its own repo)
• Upgrade BCFTools to 1.9
• Integrate with Turquoise, HMF internal monitoring system

Full release notes: https://hartwigmedical.atlassian.net/issues/?filter=10367&atlOrigin=eyJpIjoiNzc3NzAyODc3YzA0NGZhMTljNjZmODk4MDkyZTIzMDciLCJwIjoiaiJ9

v5.11

Features:

• Update Purple from 2.41 to 2.43 (see https://github.com/hartwigmedical/hmftools/releases/tag/purple-v2.43)
• Remove SAGE caller from shallow seq runs
• CRAM output enabled by default

Bugs

• Added validation of RG field in bwa invocation and improved parsing from fastq
• Fixed issue with partial data being copied on restart when CRAM step fails.

Full release notes: https://hartwigmedical.atlassian.net/issues/?filter=10365

v5.10

Features:

• Update Amber from 3.2 to 3.3
  • Improved contamination detection algorithm for shallow sequenced samples (10x).
  • See also https://github.com/hartwigmedical/hmftools/tree/master/amber
• Update GRIDSS from 2.7.2 to 2.8.3.
  • Number of bug fixes and cleanups.
  • See also https://github.com/PapenfussLab/gridss/releases
• Update Purple from 2.39 to 2.41
  • Bug fix to driver likelihoods and a new fitting deviation penalty algorithm to penalise negative copy numbers more aggressively.
  • See also https://github.com/hartwigmedical/hmftools/tree/master/purity-ploidy-estimator#version-history-and-download-links
• Format of output file with alignment now CRAM instead of BAM to decrease transfer times and storage costs
• Register CRAM output with the API at the end of the run. CRAMs file entries will now be annotated with extra data to make them easier to locate from among the other files in the run: they will be marked with a datatype (introduced separately in the API) of "reads" and associated with the sample they were created from.
• Allow pipeline users to specify input locations with JSON. In addition to the existing mechanisms (using metadata in the API or driving everything off the provided run id in the arguments), users may now provide input parameters directly. This will surely make the pipeline more accessible to third-party users.

Full Release Notes: https://hartwigmedical.atlassian.net/issues/?filter=10359

v5.9

Features:

• Argument to limit the number of lanes that will be processed concurrently during the BWA mapping stage, hence throttling the peak CPU footprint of the pipeline
• Make use of a private network mandatory for pipeline, batch and bcl2fastq
• Require customer-managed encryption keys (CMEK) for pipeline, batch and bcl2fastq
• Reduce GCP API compute engine status polling interval for the batch framework to allow larger batches to be run without exceeding global quotas on Compute Engine list and get requests.

Full Release Notes: https://hartwigmedical.atlassian.net/issues/?filter=10351

v5.8

Features:

• Add BCL conversion as a new pipeline (bcl2fastq 2.20.0.422)
• Support reading FASTQ from GCP
• Upgrade Purple from 2.34 to 2.39
• Upgrade Linx from 1.5 to 1.7
• Upgrade Amber from 2.5 to 3.2
• Add Sage v2 variant calling sub-pipeline parallel to Strelka
• Retain Sage hotspot file in output in Strelka sub-pipeline
• Option to CRAM compress pipeline output before archiving (disabled by default)

Bugs

• Include unfiltered gridss VCF in final results

Full Release Notes: https://hartwigmedical.atlassian.net/issues/?filter=10330

v5.7

• Remove the MODE argument. Always run in full pipeline mode.
• Include all resources and config in the vm image
• Version the vm image
• Fix issue with parsing read group names
• Per lane alignment logs included in final result
• Disable S3 persistence
• GRIDSS 2.7.2
• Switch to using GRIDSS driver script

Image version for deployment is pipeline5-5-7-201912121346.

Full release notes: https://hartwigmedical.atlassian.net/issues/?filter=10326&atlOrigin=eyJpIjoiYWQ5OWZiYjM1ODA4NGM0OGI0OTBlMzA0NzBhMTgxMzQiLCJwIjoiaiJ9

v5.6

Features:

• Transfer final output to GCS as well as S3
• Add SAGE hotspot VCF to final output
• Use GRIDSS jar instead of picard for viral annotation SortVcf
• Make gridss a pre-emptible job
• Include run.log and metadata in root of results for single sample
• Upgrade to JDK 11
• GRIDSS optimizations
• Reduce compute.lists requests

Bugs:

• Fixed issue with "output" folder present in results
• Fixed issue with VMs pre-empted at startup
• Fixed issue with failing single sample jobs errors not propagated to final pipeline result

Full release notes: https://hartwigmedical.atlassian.net/issues/?filter=10324&atlOrigin=eyJpIjoiNDU4MDQwZTJhZWNhNDRjMjg1MjU1YWQ5MjNjMDQ1ZmUiLCJwIjoiaiJ9

v5.5

• Upgrade of Strelka Post Process to v1.6
• Add LINX to pipeline
• Add CHORD to pipeline
• Add BACHELOR to pipeline
• Add GRIDSS "repeat masker" and "viral" annotations

https://hartwigmedical.atlassian.net/issues/?filter=10322&atlOrigin=eyJpIjoiNzhiMmFhNWI4ZjRiNGRjOWFiMTIxNWM0MDBhMmUxZmMiLCJwIjoiaiJ9