Ensembl_thesis

Contains the new healthchecks in Perl for the Ensembl database.

Healthchecks are grouped in 5 categories:

1. 'Hard' integrity: entity and referential integrity.
2. 'Soft' integrity: user-defined integrity.
3. Schema-related sanity: unique rows, blank & NULL values etc.
4. Data-related sanity: Data format & values.
5. Comparison of databases and/or database versions

A set of 'old' healthchecks will be adapted into Perl using the Ensembl API. For of the above categories a number of old healthchecks has been chosen to demonstrate the concept. Further development and optimisation will follow.

Inital set of tests to be developed: (by their old name, all can be found in: https://github.com/Ensembl/ensj-healthcheck/tree/release/83/src/org/ensembl/healthcheck/testcase/generic )

Category 1

• CoreForeignKeys
• AncestralSequencesExtraChecks - Tests on compara db - what was it doing in the generic repository?

Category 2

• AssemblyMapping
• AssemblyMultipleOverlap
• FeatureCoords
• LRG
• Meta
• ProjectXrefs
• SeqRegionCoordSystem
• SequenceLevel
• VariationDensity
• XrefTypes

Category 3

• AutoIncrement
• BlanksInsteadOfNulls
• SchemaType
• StableID

Category 4

• AssemblyNameLength
• DataFiles
• GeneCount
• NonGTACNSequency
• XrefPrefixes

Category 5

• ComparePreviousDatabases
• CoordSystemAcrossSpecies
• MySQLStorageEngine
• ProductionMeta (or another Production test)
• SeqRegionAcrossSpecies & SeqRegionAttribAcrossSpecies

(Non-ensembl modules used: Moose, Test::CSV, Getopt::Long, File::Spec)