Orchestrator V13.9.4

V13.9.4 (V13.9.3 was skipped), this ties up the final changes added to the model prior to our exit from the PRIMAGE project.

• Final recalibration

• Log more info to stdout, useful for working out suspiciously short runtimes

• BugFix: Cellularity array order was different to that used by the orchestrator

• BugFix: Incorrect value copy paste error in Primage output

• Total cell count is now conserved between successive loops and SC migration

• Add 'cell_count' to input and output files.

• If all cells die, volume ratio updated is now reported as -1.0, allowing the orchestrator to detect and rewrite to median or similar

• Add support for a calibration file referenced by the input file, to allow recalibration without recompiling the model

• Set model to only build with final supported version of FLAMEGPU (pre-reference change).

• Add default calibration file and update example input file to include new members.

Build scripts now exist for both Prometheus and Ares which can be launched using sbatch scripts/build_on_prometheus.sh or sbatch scripts/build_on_ares.sh respectively, both will produce the binary build/bin/Release/orchestrator_FGPUNB. (Note, due to library differences between machines these binaries will differ and are not compatible on the other system)

Orchestrator V13.9.2

V13.9.2

• a5b6404: Recalibrate MYCN response

The attached Prometheus binaries was compiled on PLGrid using CUDA 11.2, GCC 8.2.0 (and FGPU2 e95a261), and can be found on Prometheus at the location /net/archive/groups/plggprimage/usfd/2022-08-12_v13.9.2_fgpu2.

The attached Ares binaries was compiled on PLGrid using CUDA 11.2.2, GCC 10.3.0 (and FGPU2 e95a261). Ares does not have a shared folder that can be used for distribution.

Build scripts now exist for both Prometheus and Ares which can be launched using sbatch scripts/build_on_prometheus.sh or sbatch scripts/build_on_ares.sh respectively, both will produce the binary build/bin/Release/orchestrator_FGPUNB. (Note, due to library differences between machines these binaries will differ and are not compatible on the other system)

Orchestrator V13.9.1

V13.9.1

• 4bbfe78 Add build scripts for Uni of Sheffield A100 GPUs on Bessemer.
• 2631666 Fix CI from running out of memory
• 47f33d0 Updated calibration (see here)
• 806537c BugFix: Cellularity init was wrong at orchestrator time 0.
• 75820b9 Update PLGrid grants in scripts to plgprimage4
• 983ae02 Update code documentation
• 5843d21 Remove redundant Schwann vars
• 9636f42 Update version to 13.9.1

The attached Prometheus binaries was compiled on PLGrid using CUDA 11.2, GCC 8.2.0 (and FGPU2 620e904), and can be found on Prometheus at the location /net/archive/groups/plggprimage/usfd/2022-07-06_v13.9.1_fgpu2.

The attached Ares binaries was compiled on PLGrid using CUDA 11.2.2, GCC 10.3.0 (and FGPU2 8627ede). Ares does not have a shared folder that can be used for distribution.

Build scripts now exist for both Prometheus and Ares which can be launched using sbatch scripts/build_on_prometheus.sh or sbatch scripts/build_on_ares.sh respectively, both will produce the binary build/bin/Release/orchestrator_FGPUNB. (Note, due to library differences between machines these binaries will differ and are not compatible on the other system)

Orchestrator V13.9

V13.9

• 90210a7 Added new parameter to control neuroblast cycling and recalibrated model for parameter to control growth.

• 2835daa BugFix: The --device runtime argument was not being used.

• c034507 Cleaned up the repository and setup specific build/run scripts for Prometheus and Ares. Additionally, added a script for performing csv-to-csv parameter sweeps.

The attached Prometheus binaries was compiled on PLGrid using CUDA 11.2, GCC 8.2.0 (and FGPU2 620e904), and can be found on Prometheus at the location /net/archive/groups/plggprimage/usfd/2022-05-30_v13.9_fgpu2.

The attached Ares binaries was compiled on PLGrid using CUDA 11.2.2, GCC 10.3.0 (and FGPU2 620e904). Ares does not have a shared folder that can be used for distribution.

Build scripts now exist for both Prometheus and Ares which can be launched using sbatch scripts/build_on_prometheus.sh or sbatch scripts/build_on_ares.sh respectively, both will produce the binary build/bin/Release/orchestrator_FGPUNB. (Note, due to library differences between machines these binaries will differ and are not compatible on the other system)

Orchestrator V13.8.4

V13.8.4

• 179ccbb When cellularity is 0, report total_volume_ratio_updated as 1 (tumour volume unchanged, no cells, all matrix)

V13.8.3

• 21aa17f total_volume_ratio_updated is now reported as the relative change in internal volume.

• 21aa17f Recalibration of P_DNA_damage_pathways and P_apopchemo for the updated volume calc using LHC_Calibrate2#46

• 9c7699c BugFix: There were some issues with chemo events for runs with more than 1 chemo event (identified due to updated FGPU2).

The attached binary was compiled on PLGrid using CUDA 11.2, GCC 8.2.0 (and FGPU2 445b520), and can be found on Prometheus at the location /net/archive/groups/plggprimage/usfd/2022-04-19_v13.8.4_fgpu2.

Orchestrator V13.8.2

Two bugfixes:

• 3e9ad78 - Incorrect type used when accessing step counter during force resolution (lead to exception with latest FLAMEGPU2, previously harmless)
• 6a27d5d - The calculation of total_volume_ratio_updated was inverted.

The attached binary was compiled on PLGrid using CUDA 11.2, GCC 8.2.0 (and FGPU2 445b520), and can be found on Prometheus at the location /net/archive/groups/plggprimage/usfd/2022-04-11_v13.8.2_fgpu2.

Orchestrator V13.8.1

This is the first release build of the FLAMEGPU2 implementation of the USFD Primage model to the orchestrator.

The input/output format and runtime arguments of this FLAMEGPU2 build should match that of the FLAMEGPU1 versions (this particular branch maintains that additional interface to the model).

v13.6: This update is skipped by the orchestrated model, as it affects cellularity initialisation, which the orchestrator provides.
v13.7: Chemo damages DNA and missing DNA damage response pathways have been added.
v13.8: Altered init of telomere lengths, added missing links between chemo and Schwann cells.
v13.8.1: Update the volume calculation method and calibrate P_DNA_damage_pathways and P_apopChemo.

It additionally contains several bug fixes over the previous FLAMEGPU1 release, which were discovered whilst validating the two models were equivalent.

The attached binary was compiled on PLGrid using CUDA 11.2 and GCC 8.2.0, and can be found on Prometheus at the location /net/archive/groups/plggprimage/usfd/2022-04-07_v13.8.1_fgpu2.