1.29.11

This is a small patch release with a few critical issues being addressed.

• Use a more efficient cache strategy for raw and processed vectors.
• Require at least GROUP_AGENT authority for adding update audit events
• Fix search for deprecated/removed ontology terms
• Allow groups to be updated in the user manager interface

1.29.6

This is a small patch release that backports a number of fixes from the 1.30 series.

• Fix ArrayDesign proxy class leaking in search results (a more definitive fix is scheduled for the 1.30 series)
• Fix additivity for Jersey logging to prevent annoying logs to reach Slack
• Fix numerous potential NPEs with loadValueObject and loadValueObjectById
• Cancel ontology loading threads when the OntologyService is disposed
• Move experiment-related routines from AnalysisDao into SingleExperimentAnalysisDao and make those type-safe
• Update dependencies

1.29.5

Tag hotfix

1.29.3

• use native Hibernate enum types, eliminating about 4 KLOC of boilerplate
• fix tons of potential NullPointerExceptions and produce useful error messages when an entity cannot be loaded
• fix DWR converter for SequenceTypeValueObject
• use Log4j JUL and JCL artifacts to intercept more logs
• fix NPE when biomaterials cannot be ordered properly

1.29.2

• Fix a few missed queries when transitioning to SessionFactory APIs
• add support for filtering platforms by taxon ID, common name, etc.
• fix AffyScanDateExtractorTest when running the test suite outside America/Vancouver TZ, parsing was correct, but the test was using the system TZ
• enforce JavaBeans contract to all our VOs which fixes missing default constructor in TaxonValueObject, there's a new validator that runs in the test suite
• fix incorrect result type for phenotype associations
• fix negative limit behaviour in dataset manager
• improve resolution of properties in getObjectFilter() and getSort() in FilteringService

1.29.1

Fix a few regressions from the 1.29 release.

• add missing @Transactional annotations on CompositeSequenceServiceImpl
• fix incorrect setParameter when thawing mulitple BLAT results (should have been setParameterList)
• translate taxon ID -1 to null in dataset controller, since the frontend uses that for representing no taxon

1.29.0

Changeset

• Update to Hibernate 4
• Keep track of genome, gene annotations, and other external databases
• Improve auditing by performing it using a before advice (auditing now happens before the entity is saved)
• Improve pointcuts
• Make the retry policy more targeted
• Update pavlab-starter-parent to 1.2.4
• Use int for offset/limit throughout the codebase
• Relocate Gemma CLI under its own module
• Make Gemma CLI commands injectable
• Replace Ehcache usages with Spring Cache wherever possible

Highlights

Keep track of external databases updates

Gemma models external source of data like genomes and gene annotations via the ExternalDatabase model. This update makes it auditable and introduce a new interface for versioned entities. This will allow us to attach version metadata and keep records using the audit trail when these change over time.

New external databases have also been inserted in the database to model genome annotations and annotations used by the RNA-Seq pipeline.

Update Hibernate to 4.2.21

This update is paving the way to update Gemma to Spring 4 #508 and subsequently Spring 5 #116 as it eases the transition to Hibernate 5. We got rid of all the HibernateTemplate usages and other Hibernate 3-specific code such as org.hibernate.classic.* APIs.

In addition, we also switched from Session.save to Session.persist and introduced support for Session.merge. This will allow us to eventually solely use standard JPA routines.

Spring Cache

Favour Spring Cache over direct access to Ehcache API. We currently only use the Ehcache API for displaying hit/miss statistics in the admin section of Gemma Web. All custom-defined caches were moved to the ehcache.xml configuration.

1.28.2

This patch release introduces new endpoints for retrieving expression data by quantitation type.

• fix serialization of ArrayDesignValueObject by making its associated taxon serializable too
• improve implementation of ArrayDesignController.downloadAnnotationFile() and report which platform/file cannot be generated
• update pavlab-starter-parent to 1.1.8
• annotate all possible search result types in the OpenAPI specification
• add a link to the QT download when available

Download quantitations at a glance from the "Quantitation Types" tab

When the new API endpoint to download expression data by quantitation type, we can now create links in the administrative section of Gemma to download specific expression vectors.

1.28.1

This is a small release that fixes a few issues with the RESTful API specification.

• restore objectClass visibility in AnnotationValueObject
• fix incorrect response types for annotations search endpoints returning datasets

1.28.0

This minor release of Gemma is the culmination of several months of work.

Changeset

• improved Search API available through the RESTful API
• dbcp2 for pooling UCSC Genome Annotation Database connections
• new exception hierarchy for filtering data with FilteringException
• major code cleanups
• removal of distributed Ehcache, JMS, etc.
• dependency update
• dependency management with pavlab-starter-parent
• RESTful API overhaul and cleanup
• introduction of non-null annotations
• search result scoring

Highlights

New Search API

We've revamped the search API and exposed in the Gemma RESTful API.

Refocus of Gemma on differential gene expression analysis

Various features were hidden from Gemma Web interface as we focus Gemma on differential expression analysis of genes.

Major overhaul of the Gemma RESTful API

We went through our whole RESTful API to sweep some dust, hide unneeded features, fill missing attributes. We renamed a few things to make it more consistent, keeping older names whenever possible for backward-compatibility.

The API version has been bumped to 2.5.0.

Head over https://gemma.msl.ubc.ca/resources/restapidocs/ for the new interactive docs based on Swagger UI.

Major code cleanups

We removed the support for the distributed cache Terracotta and JMS. This had a drastic impact on the amount of LOCs that need maintenance.