tierup

tierup reanalyses Tier 3 variants in undiagnosed GeL rare disease cases.

tierup answers the following questions derived from the GeL tiering rules:

• Is the variant in a green gene in a panel app panel assigned to the case?

• Does the variant mode of inheritance match the gene's in the panel app panel?

Tier 3 variants that meet these criteria are labelled 'tier_1' or 'tier_2' in the tier_tierup field of the tierup output CSV.

Installation

pip install jellypy-tierup

Guides

Reanalyse Tier 3 variants in a GeL rare disease case

1. Create a configuration file with CIPAPI details. For example:

    # config.ini
    [pyCIPAPI]
    client_id = YOUR-GEL-CLIENT-ID
    client_secret = YOUR-GEL-CLIENT-SECRET

2. Run tierup For example, interpretation request 1234 version 2 could be analysed with:

    tierup --irid 1234 --irversion 1 --config config.ini

   If the intepretation request data is available locally in json format, then you can pass the file directly:

    tierup -j interpretation_request.json --config config.ini

3. View results

   • *.tierup.csv - The tier_tierup column in the results file contains the new variant tier determined by tierup. Each row is a report event for a variant in the proband. Note: The same variant may have multiple report events depending on the number of assigned gene panels, mode of inheritance and penetrance models analysed.

TierUp output fields (*.tierup.csv)

Field	Description
interpretation_request_id	Identifier for GeL case
tier_tierup	Tiering result from tierup: tier_3_not_in_panel, tier_3_red_or_amber, tier_3_green_moi_mismatch, tier_2, tier_1
tier_gel	Initial GeL tier. Previously tier field in tierup 0.2.0
assembly	Reference genome
chromosome	Chromosome number
position	Genomic coordinate
reference	Reference allele
alternate	Proband Alternate allele
consequences	Sequence ontology terms for the variant consequences in all relevant transcripts
zygosity	Zygosity in the proband
segregation	Pattern calculated using genotypes from family members
penetrance	Penetrance model used to assess the variant
tiering_moi	Mode of inheritance for the variant in the proband. Previously inheritance field in tierup 0.2.0
tu_panel_hash	Panel unique hash id used in TierUp analysis
tu_panel_name	Panel name used in TierUp analysis
tu_panel_version	Panel version used in TierUp analysis
tu_panel_number	Panel id used in TierUp analysis
tu_panel_created	The date the panel version used in TierUp analysis was created
tu_run_time	Time TierUp was initiated
pa_ensembl	Ensembl identifier for the gene in panelapp panel
pa_hgnc_id	HGNC ID from the panel app panel
pa_gene	Gene symbol from the panel app panel
pa_moi	Mode of inheritance for the gene in the panelapp panel version
pa_confidence	Current gene confidence level from the latest version of the panel app panel: 4 or 3 = Green; 2 = Amber; 1 or 0 = Red; Green genes indicate TieredUp variant.
extra_panels	Panel ID pointers updated via TierUp check for depreciated Panels in GeL directory
re_id	Unique identifier for the report event. Previously event_id field in tierup 0.2.0
re_panel_id	Report event panel id
re_panel_version	Report event panel version
re_panel_source	Report event panel source. All expected to be panelapp
re_panel_name	Report event panel name
re_gene	Report event gene symbol
justification	GeL eventJustification field which describes why the variant was tiered
created_at	Creation date for the GeL interpreted genome was created
software_versions	Software versions used in the GeL analysis
reference_db_versions	Reference database versions used int he GeL analysis
tu_version	jellypy-tierup version used to generate results

Constraints

• tierup input is limited to undiagnosed rare disease cases. The tool therefore raises an error if users attempt to process a solved case.

• Access to case data from GeL is only possible on the HSCN

• Interpretation request data must comply with the GeL v6 interpretation request model

• tierup does not analyse CNVs

• tierup does not account for information on variant penetrance

Support

Please raise issues, message the #jellypy slack channel or send an email.