v2.9.1 "Amtrak"

In a major change to the previous versions, the 2.9.1 release also removes all of the byconServices apps and libraries - now refactored in https://codeberg.org/Progenetix/byconservices (additionally to the previous library move & rewrites to https://codeberg.org/Progenetix/byconplus). Still no full release; more work on documentation, settings/configuration & deployment methods needed.

v2.9.0 "Foggy Bottom"

The v2.9.0 "Foggy Bottom" release is a preparation for a v3 which will include "settled" Beacon v3 features as well as a complete separation from services (now still included).

v2.8.0 "genemaps"

This release contains changes focussed on data aggregations and gene hit statistics, as well as many areas of restructured code. For details please see thechanges tracker.

v2.7.0 "That Was 2025"

This release chiefly introduces data summaries ("Aggregations") as a full stack WiP development; from data aggregation in MongoDB over a summary response format proposal to JS dashboards with 1D and 2D summary data displays (e.g. stacked bar charts).

2025-12-31 (v2.7.0 "That Was 2025")

• beaconplusWeb front end updates
  • refactored data summaries source in src/components/summaries/
  • improving of documentation in https://bycon.progenetix.org/data-summaries/

2025-12-22 (v2.7.0+pre1)

• de-nested the classificationTrees into single tupe specific files
  and moved the additional sources for their generation outside of the project
  • also renamed some files/filter types (e.g. NCIT ==> NCITneoplasm, NCITsex ==> pgxSex)
• in beaconplusWeb changes to AggregationPlots
  • moved from Victory to Plotly; this requires npm install react-plotly.js plotly.js
• new ___BEACON_ROOT___/beacon/aggregation_terms endpoint to indicate concepts
  for available data summaries

2025-12-12 (v2.6.9)

The recent changes addressed several issues with the aggregation implementation

• now ByconSummaries class
  • dimension agnostic, i.e. supports 1D, 2D, ... nD aggregations through the
    same methods
  • TODO: So far cross-entity aggregations are supported through some workarounds;
    e.g. some information stored in individual (sex, age, followup...) is
    repeated in the biosamples' individual_info property to allow biosample
    aggregations on individual properties. A more generic solution would be
    desirable, e.g. utilizing MongoDB's $lookup.
• added (expanding ...) doumentation page about the Beacon Aggregation development

2025-11-28 (v2.6.8 "Aggregator")

• extending aggregation responses to support 2-dimensional aggregations
  • implemented in ByconSummaries class
  • now also supports plot_type: "stackedBar" for the dashboard

2025-11-18 (v2.6.7 "DashboardDabbler")

• implemented a basic dashboard at ___BEACON_ROOT___/dataDashboard/ using the
  draft version of a Beacon v2.n aggregation response
  • simple Victory based bar plots defined in beaconplusWeb/src/components/AggregatedPlots.js
  • only one-dimensional so far; 2-dimensional for stacked plots coming up
• ... based on the ByconSummaries class to handle aggregation definitions
  and processing
  • added aggregation_definitions.yaml file for defining available aggregations
    and their parameters
  • modified ByconResultSets to include aggregation processing if requested
    through the aggregated granularity parameter
• added a new individual_info object property to biosamples which contains
  data extracted from the corresponding individual, to allow easier aggregation
  • one has to run the updated housekeeping.py script to populate this field

  individual_info: {
    index_disease: {
      disease_code: { id: 'NCIT:C8851', label: 'Diffuse Large B-Cell Lymphoma' },
      stage: { id: 'NCIT:C27966', label: 'Stage I' },
      clinical_tnm_finding: [ { id: 'NCIT:C48720', label: 'T1 Stage Finding' } ],
      followup_state: { id: 'EFO:0030049', label: 'dead (follow-up status)' },
      followup_time: 'P1M',
      onset: { age: 'P41Y', age_days: 14974 },
      followup_days: 30
    },
    sex: { id: 'NCIT:C16576', label: 'female' }

v2.6.6 "GlobeProjector"

2025-11-04 (v2.6.6 "GlobeProjector")

• created ByconHO class in beacon_responses.py and deleted handover_generation.py
• created new /services/sampleglobe endpoint for plotting sample provenance
  on a globe projection using Globe.gl

2025-11-03 (v2.6.5 "geonames")

• updated the geolocs database with the latest GeoNames dump (2025-10-29)
  • now using cities500, i.e. 10x the number of previous entries
  • created geoUpdater.py housekeeping script for future updates of the
    geolocs collection
  • added a function to housekeeping.py to update the geo_location in
    biosamples from the existing long, lat data to pull the nearest match from
    the geolocs collection
    • side effect are now some rather granular annotations; e.g. New York
      coordinates which point to Manhattan are now labeled as "Financial District"
      instead of "New York City"
    • unknown coords ar now all mapped to Atlantis whish - as everybody knows -
      sits below the Bermuda Triangle (centered at about lat: 25, long: -71)
  • ... above now handled by ByconGeoResource class

v2.6.4 "___BEACON_ROOT___"

• changed from hard coded https://progenetix.org to template ___BEACON_ROOT___
  in map, schema and configuration files and for handover generation
  • the ___BEACON_ROOT___ is replaced during request processing through the
    detected f"{REQUEST_SCHEME}://{HTTP_HOST}"
• fixed UCSC link generation for BeaconPlus
• removed beaconMap.yaml from schemas, since now part of config/beacon_map.yaml
• renamed ENV to HTTP_HOST for consistency
• new dict_replace_values function in bycon_helpers.py

... and from 2.6.3

• schema file restructuring
  • removed all JSON schemas fromn the source tree since they are generated
    from YAML anyway and changed the file targets in schema_parsing.py to
    .yaml accordingly
  • flattened schema hierarchy structure since there is no need to distinguish
    between src and json directories anymore
  • removed the schemas/bin directory since no json files are generated anymore
  • adjusted documentation generation in markdowner.py accordingly
• cleaned out code w/ removal of "pgxVariant" references since using vrsVariant
  • e.g. in ByconVariant and ByconBundler

v2.6.3 "Small Surgery"

This release

• gets rid of JSON schema file since they are processed from YAML source anyway; and rearranges files in a slightly flattened out structure
• restructures configration files, so that e.g. paths are now matched from the map, not as separate parameters; and generally removes some of the custom properties of the configurations
• removes some parameter processing functions / a file and adds more hard coded load to the root __init__.py
• restructures e.g. database related configs into BYC_DBS
• removes the pgxVariant format (now passing everything through "VRS + legacy properties"

v2.5.0 "Forked"

Please see the version tracker at https://bycon.progenetix.org/changes/.

v2.4.9 "Montréal"

2.4.9 cleanup

v2.4.8 "Friday the 13th"

Please see https://bycon.progenetix.org/changes/