search

fractal-analytics-platform / fractal-client

Command-line client for Fractal
https://fractal-analytics-platform.github.io/fractal-client
BSD 3-Clause "New" or "Revised" License
45 stars 1 forks source link

Overview Test Datasets #213

Closed jluethi closed 10 months ago

jluethi commented 2 years ago

This is just a collection of tests for reference, not really an issue. Let's use it to discuss when we need more tests or when there is an issue with test data.

Grid-based

1) UZH path: /data/active/fractal/3D/PelkmansLab/CardiacMultiplexing/Cycle1_testSubset FMI path: BASEPATH/Fractal_Dev/test_datasets/UZH/3D_CardiacMultiplexing/Cycle1_testSubset (maybe slightly old metadata, get fresh from Zenodo) Image count: 120 Metadata: Yes, correct metadata Has been processed: Yes, Fractal 0.1.1

Description: Core test set, also on zenodo: https://zenodo.org/record/7057076 2x2 sites in a single well, 10 z planes, 3 channels

2) /data/active/fractal/3D/PelkmansLab/CardiacMultiplexing/Cycle1_tinyTestSet Image count: 4 Metadata: Yes Has been processed: Yes, Fractal 0.1.1

Description: Tiny test data with real images. 2 sites, 2 z planes, 1 channel Also on Zenodo: https://zenodo.org/record/7059515

3) /data/active/fractal/3D/PelkmansLab/CardiacMultiplexing/Cycle1_5x5_10wells Image count: 24900 Metadata: No Has been processed: Fractal monolith prototype

Description: 10 wells with 25 FOVs, 3 channels. Varying number of Z planes between channels All FOVs are a 5x5 grid (5 rows, 5 cols), horizontal arrangement.

4) /data/active/fractal/3D/PelkmansLab/CardiacMultiplexing/Cycle1_5x5_10wells_constantZ Image count: 14252 Metadata: Yes Has been processed: Fractal monolith prototype (without using the metadata)

Description: 10 wells with 25 FOVs, 3 channels. Like 3), but with constant number of Z planes (only the first 19 for each FOV).

5) UZH path: /data/active/fractal/3D/PelkmansLab/CardiacMultiplexing/Cycle1_9x8_singleWell FMI path: BASEPATH/Fractal_Dev/test_datasets/UZH/3D_CardiacMultiplexing/Cycle1 Has been processed: Fractal monolith prototype (without using the metadata) Image count: 4104 Metadata: Yes

Description: Large, single well test set (too large to fit into napari memory without pyramids). 9x8 well, 3 channels, 19 z planes

6) UZH path: NONE FMI path: BASEPATH/Fractal_Dev/test_datasets/FMI/grided_overlap_184planes_100FOVs Image count: 36800 Metadata: Yes, correct metadata Has been processed: No

Description: Organoid data with 50 pixel overlap. 10x10 sites in a single well, 184 z planes, 4 channels

7) UZH path: /data/active/fractal/3D/PelkmansLab/plate_varying_wells FMI path: NONE Image count: 132 Metadata: Yes, synthetic but adapted correctly Has been processed: Yes

Description: Small test setup for a plate with varying well sizes (in xy & in Z)

2D only

UZH path: /data/active/fractal/2D/hiPSC_Slice/2D_test_set Has been processed: Image count: 216 Metadata: Yes

Description: 2D only dataset, synthetically created, just an individual Z planes, renamed images with fitting metadata.

Multiplate

1) /data/active/fractal/3D/PelkmansLab/CardiacMultiplexing/Multiplate_2x2_singleWell Image count: 2x 120 Metadata: Yes Has been processed: No

Description: 2 plates, each containing the data from Grid-based 1) test in separate folders. Plate 2 well has been renamed to B05.

Multiplex

1) /data/active/fractal/3D/PelkmansLab/CardiacMultiplexing/Multiplex_2x2_singleWell Image count: 2x 120 Metadata: Yes Has been processed: No

Description: 2 cycles, each containing the same image data from Grid-based 1).

2) UZH path: /data/active/fractal/3D/PelkmansLab/CardiacMultiplexing/tiny_multiplexing FMI path: TBD Image count: 48 Metadata: Yes

Description: Synthetic multiplexing test data for Fractal from the 20200810-CardiomyocyteDifferentiation14 from Joel Lüthi Contains 2 wells, B03 & B05. Only 2 FOV, 2 Z slices each (slices 5 & 6 of the stack, renamed to 1 & 2) + Synthetic metadata files for each cycle

Cycle 1: Only contains DAPI (C01) Cycle 2: Contains DAPI (C01) & Na/K ATPase (C03) Cycle 3: Contains DAPI (C01), HSP60 (C02) & bTubulin

Search First

1) UZH path: /data/active/fractal/Liberali/1_well_15_fields_20_planes_SF_w_errors/D10_R1/220304_172545_220304_175557 FMI path: BASEPATH/Fractal_Dev/test_datasets/FMI/1_grid_SF/1_well_15_fields_20_planes_SF_w_errors/HCA2_2_D10_R1/220304_172545_220304_175557 Image count: 836 Metadata: Yes

Description: Search first example data. 4 channels (different channels than the default!) 1 well, 15 FOVs (but some missing => 11 present, FOV numbering is not continuous) FOV placement is in the metadata, it's a grid-based search first experiment.

Caution: Folder contains extra TIF files from the microscope that should not be parsed. Examples: DC_sCMOS #2_CAM2.tif SC_BP445-45_4x_M01_CH01.tif (The Yokogawa generates those files, Fractal should ignore them)

2) /data/active/fractal/3D/PelkmansLab/ZebrafishMultiplexing/cycle0 Image count: 18526 Metadata: Yes

Description: Large search-first zebrafish dataset. 4 wells, 3 channels. 247 Z planes (=> good test for some memory scaling behaviors) => Let's test this once search-first is running

Also has a second cycle in /data/active/fractal/3D/PelkmansLab/ZebrafishMultiplexing/cycle1 once multiplexing is running

3) /data/active/fractal/3D/PelkmansLab/ZebrafishMultiplexingAdvanced

Complex version of test case 1). Partial cycles 2 & 3 (both split into 2 folders) => tackle later to test https://github.com/fractal-analytics-platform/fractal-tasks-core/issues/35

4) UZH path: /data/active/fractal/Liberali/FractalTesting20220124/gridless_Yokogawa_recording-FMI/20220316_sec_FOCM_test-R1_E2/day3/TIF FMI path: BASEPATH/Fractal_Dev/test_datasets/FMI/2_gridless_SF/20220316_sec_FOCM_test-R1_E2/day3/TIF (but metadata is currently in the META folder, same level as TIF => @gusqgm can you put a copy in the TIF folder so Fractal would work here as well?) Image count: 4606 Metadata: Yes

Description: Gridless dataset from Nicole. Good test set to check the gridless search first that is smaller than 2) 49 FOVs, 94 Z planes. Single channel

5) UZH path: data/active/fractal/Liberali/1_well_16_fields_23_planes_SF/201025MK001EB FMI path: BASEPATH/Fractal_Dev/test_datasets/FMI/1_grid_SF/1_well_16_fields_23_planes_SF/201025MK001EB Image count: 1104 Metadata: Yes

Description: Grid Search First like Search-First 1), but without errors in the metadata.

6) UZH path: /data/active/fractal/Liberali/FractalTesting20220124/210305NAR005AAN FMI path: BASEPATH/Fractal_Dev/test_datasets/FMI/1_grid_SF/2_well_20_fields_84_planes_dense/210305NAR005AAN Image count: 13440 Metadata: Yes

Description: FMI dataset, 2 wells, 20 FOVs, 84 Z planes

7) UZH path: /data/active/fractal/Liberali/20221103_SFtestCV8000/CV8K_images FMI path: TBD Image count: 66 Metadata: Yes (original from the microscope)

Description: CV8K test set from Nicole. 4 FOVs at different positions in a well, search first. Successfully processed with 1.0a10 server pre-release, but still an issue with the measurement

Time

1) /data/active/fractal/3D/PelkmansLab/Adrian_Time/20220414_hiPSC_MS2_1well_10sites_80tp

Time test case from Adrian with 80 timepoints, 25600 files. To be tested later

Large datasets

1) UZH path: /data/active/jluethi/20200810-CardiomyocyteDifferentiation14/Cycle1/images_renamed FMI path: Not available Image count: 163512 Metadata: Yes

Description: 23 wells, 72 FOV each. 3 channels. ~700GB of 3D image data. Only available on UZH side.

2) UZH path: Not available FMI Path: 201025_003_05_MediumScreen Only available on the FMI side, big dataset. 171 wells à 20 fields, 23 planes per field, 3 channels

Many Z planes

1) UZH path: /data/active/fractal/Liberali/3_manyZplanes_20220507GCPLEX_R0 FMI path: BASEPATH/Fractal_Dev/test_datasets/FMI/3_manyZplanes_20220507GCPLEX_R0 Image count: 117 Metadata: Yes (subset of the full metadata, modified to fit the test set)

Description: Single FOV & 1 channel of a large dataset with 117 Z planes to test correct parsing of Z levels.

jluethi commented 2 years ago

I updated the test datasets with paths on the FMI side and with additional search-first datasets from the FMI

gusqgm commented 2 years ago

New addition: grided FOV with overlap.

Grid-based

6) UZH path: NONE FMI path: BASEPATH/Fractal_Dev/test_datasets/FMI/grided_overlap_184planes_100FOVs Image count: 36800 Metadata: Yes, correct metadata Has been processed: No

Description: Organoid data with 50 pixel overlap. 10x10 sites in a single well, 184 z planes, 4 channels

jluethi commented 1 year ago

@tcompa I added a new multiplexing test dataset. It's tiny, but covers some of the complexities. Total size is 300 Mb.

UZH path: /data/active/fractal/3D/PelkmansLab/CardiacMultiplexing/tiny_multiplexing FMI path: TBD Image count: 48 Metadata: Yes

Description: Synthetic multiplexing test data for Fractal from the 20200810-CardiomyocyteDifferentiation14 from Joel Lüthi Contains 2 wells, B03 & B05. Only 2 FOV, 2 Z slices each (slices 5 & 6 of the stack, renamed to 1 & 2) + Synthetic metadata files for each cycle

Cycle 1: Only contains DAPI (C01) Cycle 2: Contains DAPI (C01) & Na/K ATPase (C03) Cycle 3: Contains DAPI (C01), HSP60 (C02) & bTubulin

jluethi commented 1 year ago

I added:

Many Z planes

1) UZH path: /data/active/fractal/Liberali/3_manyZplanes_20220507GCPLEX_R0 FMI path: BASEPATH/Fractal_Dev/test_datasets/FMI/3_manyZplanes_20220507GCPLEX_R0 Image count: 117 Metadata: Yes (subset of the full metadata, modified to fit the test set)

Description: Single FOV & 1 channel of a large dataset with 117 Z planes to test correct parsing of Z levels.

jluethi commented 10 months ago

This has been made obsolete by the fractal demos repo, which contains actual workflows for test datasets: https://github.com/fractal-analytics-platform/fractal-demos