[Run issue]: DBSCAN results

[ShahinGHD]

Contact Details

shahin.ghasemi@gmail.com

What was the issue?

Hi everyone.

I am having problems understanding the DBSCAN results.
I have two TsSphericalCell with 20um diameter and 8um diameter nuclei. these two cells are attached so the center-to-center distance is 20 um.
I put 1000 Ac-225 sources in one cell (call it the source cell) with 10000 histories, and I am trying to score the double-strand breaks in the other cell that has no activity in it.
I am also using : "g4em-livermore" "g4decay" "g4radioactivedecay" "g4em-dna" physics

Here are the questions I have:

1- the number of DSBs in the cell with 1000 sources and 10000 histories is 3530000 in total (Complex DSB not included). Then, at the next adjacent cell, the DSB number drops to 331. I know there is an inverse square law effect here but does it make sense that it drops by this magnitude only 20um away? what is the justification?

2- I am getting different results when putting "TsEMDNAPhysics" and "g4em-dna". Which one is the correct one to use in this case? the "g4em-dna" showing almost double the number in SSBs but the DSBs are the same!

3- Based on the geometry described above, do you recommend using step by step scorer or sticking with DBSCAN to measure the DNA breaks at different distances from the "source cell"?

Thanks
Shahin

TOPAS-nBio category

Scoring

Version

3.0

What Operative System (OS) are you running on?

Windows (WSL)

Relevant log output

#############
# GEOMETRY #
#############

# WORLD
d:Ge/World/HLX			= 1500 um
d:Ge/World/HLY			= 1500 um
d:Ge/World/HLZ			= 1500 um
s:Ge/World/Material			= "G4_WATER"
s:Ge/World/Color            = "lightBlue"

# WATER PHANTOM
s:Ge/WaterPhantom/Type		= "TsBox"
s:Ge/WaterPhantom/Parent	= "World"
s:Ge/WaterPhantom/Material	= "G4_WATER"
dc:Ge/WaterPhantom/HLX		= 1500 um
dc:Ge/WaterPhantom/HLY		= 1500 um
dc:Ge/WaterPhantom/HLZ		= 1500 um


b:Ge/QuitIfOverlapDetected  = "False"

## Spehrical Cells for IGPC patients
s:Ge/SourceCell/Type="TsSphericalCell"
d:Ge/SourceCell/CellRadius = 10 um
s:Ge/SourceCell/Parent = "WaterPhantom"
s:Ge/SourceCell/Material = "G4_WATER"
d:Ge/SourceCell/Nucleus/NucleusRadius= 4. um
s:Ge/SourceCell/Nucleus/Material="G4_WATER"
s:Ge/SourceCell/Nucleus/Color="White"
s:Ge/SourceCell/Nucleus/DrawingStyle="solid"

#s:Ge/MyCell1/Type="TsSphericalCell"
s:Ge/MyCell1/Type="TsSphericalCell"
d:Ge/MyCell1/CellRadius = 10 um
s:Ge/MyCell1/Parent = "WaterPhantom"
s:Ge/MyCell1/Material = "G4_WATER"
d:Ge/MyCell1/Nucleus/NucleusRadius= 4. um
s:Ge/MyCell1/Nucleus/Material="G4_WATER"
s:Ge/MyCell1/Nucleus/Color="Red"
s:Ge/MyCell1/Nucleus/DrawingStyle="solid"
d:Ge/MyCell1/TransZ              = 20 um

##########
# SOURCE #
##########
s:So/RadionuclideSource/Type                       = "Distributed"
s:So/RadionuclideSource/Component                  = "SourceCell"
s:So/RadionuclideSource/BeamParticle               = "GenericIon(89, 225)" #Ac-225
d:So/RadionuclideSource/BeamEnergy                 = 0. MeV
ic:So/RadionuclideSource/NumberOfHistoriesInRun    = 10000
ic:So/RadionuclideSource/NumberOfSourcePoints      = 1000
bc:So/RadionuclideSource/RedistributePointsOnNewRun = "False"

sv:Ph/Default/Modules       = 4 "g4em-livermore" "g4decay" "g4radioactivedecay" "g4em-dna"

s:Sc/DBSCAN/Quantity = "DBSCAN"
s:Sc/DBSCAN/Component  = "SourceCell"
s:Sc/DBSCAN/OutputFile = "Ac225__1000Source_10000Histories_0umAway"
s:Sc/DBSCAN/OutputType = "ROOT"
b:Sc/DBSCAN/OutputToConsole = "True"
s:Sc/DBSCAN/IfOutputFileAlreadyExists = "Increment"
# DBSCAN configuration parameters
d:Sc/DBSCAN/MinimumDistanceForDSB       = 3.2 nm  # Distance between SSB to form a DSB ~10 bp
u:Sc/DBSCAN/SampleHitsWithProbability   = 0.16    # Hits are scorer with 16% probability
d:Sc/DBSCAN/LowerEnergyForSamplingSSB   = 5.0 eV  # Hits are sampled from a ramp distribution between
d:Sc/DBSCAN/UpperEnergyForSamplingSSB   = 37.5 eV # these two values: 5.0 eV -> 0.0; 37.5 eV -> 1.0
i:Sc/DBSCAN/MinimumNumberofSSBtoFormDSB = 2       # Minimum number of SSB to form a DSB

s:Sc/DBSCAN1/Quantity = "DBSCAN"
s:Sc/DBSCAN1/Component  = "MyCell1"
s:Sc/DBSCAN1/OutputFile = "Ac225_1000Source_10000Histories_20umAway"
s:Sc/DBSCAN1/OutputType = "ROOT"
b:Sc/DBSCAN1/OutputToConsole = "True"
s:Sc/DBSCAN1/IfOutputFileAlreadyExists = "Increment"
# DBSCAN configuration parameters
d:Sc/DBSCAN1/MinimumDistanceForDSB       = 3.2 nm  # Distance between SSB to form a DSB ~10 bp
u:Sc/DBSCAN1/SampleHitsWithProbability   = 0.16    # Hits are scorer with 16% probability
d:Sc/DBSCAN1/LowerEnergyForSamplingSSB   = 5.0 eV  # Hits are sampled from a ramp distribution between
d:Sc/DBSCAN1/UpperEnergyForSamplingSSB   = 37.5 eV # these two values: 5.0 eV -> 0.0; 37.5 eV -> 1.0
i:Sc/DBSCAN1/MinimumNumberofSSBtoFormDSB = 2       # Minimum number of SSB to form a DSB
i:Ts/Seed = 651
i:Ts/NumberOfThreads                         = 0

Code of Conduct

• I agree to follow this project's Code of Conduct

Attestation

• I testify that I have fully read the bug reporting guideline beforehand.

[JoseRamosMendez]

Hi there!

A few things. Do not mix electromagnetic models (Livermore and g4em-dna) in the same physics list. If you want to use geant4-dna in the cells and Livermore outside the cell (for efficiency purposes), then you can try physics per regions as shown here. Something like

s:Ge/SourceCell/AssignToRegionNamed = "geant4dna"
s:Ge/MyCell1/AssignToRegionNamed = "geant4dna"

sv:Ph/Default/Modules       = 3 "g4em-livermore" "g4decay" "g4radioactivedecay"
s:Ph/Default/ForRegion/geant4dna/ActiveG4EmModelFromModule = "g4em-dna"

Also, in this case, you could build the cell geometry by hand to score with DBSCAN only in the cell nucleus rather than the whole cell. I guess that's why you're getting a kind of large number of DBS.

Regarding TsEmDNAPhysics vs. geant4-dna, it could be the different lower energy limits for ionization and elastic scattering models that would produce a different spatial distribution of ionization (hence SSB) events. The choice is up to you and how you interpret simulation when comparing with measured data.

Nevertheless, only geant4-dna is compatible with the physics per regions as shown above.

Best!

[ShahinGHD]

Thank you so much for the detailed answer.