GPL-770-1 Update the get cherrypicked samples queries for the new Beckman events about lighthouse

[KatyTaylor]

Test this by running the positives report from lighthouse ui, and making sure the contents of the 'LIMS submission' column is correct:

• It should say 'Yes' for any samples that have been cherrypicked the Sentinel way ('cherrypick_layout_set' event)
• It should say 'Yes' for any samples that have been cherrypicked the Beckman way ('lh_beckman_cp_destination_created' event)
• It should say 'No' for other samples

[Chris-Friend]

Joins on sample UUID. Looks for the Beckman destination plate creation event type for a specific sample

SELECT mlwh_sample.description AS `root_sample_id`, mlwh_lh_sample.plate_barcode AS `plate_barcode`, 
mlwh_sample.phenotype AS `Result_lower`, mlwh_lh_sample.coordinate AS `coordinate`
FROM mlwhd_mlwarehouse_devdata.sample AS mlwh_sample
JOIN mlwhd_mlwarehouse_devdata.lighthouse_sample AS mlwh_lh_sample ON (mlwh_sample.uuid_sample_lims = mlwh_lh_sample.lh_sample_uuid)
JOIN mlwhd_mlwh_events_proddata.subjects AS mlwh_events_subjects ON (mlwh_events_subjects.uuid = UNHEX(REPLACE(mlwh_lh_sample.lh_sample_uuid, "-", "")))
JOIN mlwhd_mlwh_events_proddata.roles AS mlwh_events_roles ON (mlwh_events_roles.subject_id = mlwh_events_subjects.id)
JOIN mlwhd_mlwh_events_proddata.events AS mlwh_events_events ON (mlwh_events_events.id = mlwh_events_roles.event_id)
JOIN mlwhd_mlwh_events_proddata.event_types AS mlwh_events_event_types ON (mlwh_events_event_types.id = mlwh_events_events.event_type_id)
-- f" WHERE mlwh_sample.description IN %(root_sample_ids)s"
-- f" AND mlwh_lh_sample.plate_barcode IN %(plate_barcodes)s"
AND mlwh_events_event_types.key = "lh_beckman_cp_destination_created"
GROUP BY mlwh_sample.description, mlwh_lh_sample.plate_barcode, mlwh_sample.phenotype, mlwh_lh_sample.coordinate;

[harrietc52]

Do we need to back populate the lh_sample_uuid in the lighthouse_samples table, as 19/22 million records do not have a lh_sample_uuid value, therefore the JOIN may not work as intended? Could some how the combo key root_sample_id / rna_id / result be used to identify the correct sample uuid

[Chris-Friend]

Do we need to back populate the lh_sample_uuid in the lighthouse_samples table, as 19/22 million records do not have a lh_sample_uuid value, therefore the JOIN may not work as intended? Could some how the combo key root_sample_id / rna_id / result be used to identify the correct sample uuid

Crawler's update_mlwh_and_dart_with_legacy_samples migration should handle this. It needs to be run before a sample can be added to DART (be processed by the Beckman workflow), and gives samples lh_sample_uuids