Onetahi

Use of pgSIT technology on the Onetahi island of the Tetiaroa atoll in French Polynesia.

Data sources and sims:

• MGDrivE sim (scripts to generate datasets)
• Transforming insect population control with precision guided sterile males with demonstration in flies
• Reply to ‘Concerns about the feasibility of using “precision guided sterile males” to control insects’

Experiment Nomenclature

Folders and files follow this naming convention:

• E_pop_ren_res_mad_mat
  • pop: Population size (male and female) per node (x1)
  • ren: Number of weekly releases (x1)
  • res: Release size (fraction of the stable population x100)
  • mad: Adult lifespan reduction (x100)
  • mat: Male mating reduction (x100)

For the breakdown of the AOI sets, look at the pgSIT gene definition.

Exported metrics (MOI) for the drive are:

• POE: Probability of elimination
• WOP: Total sum of time below the threshold
• TTI: First break below the threshold
• TTO: Last break below the threshold
• RAP: Fraction of the population with the genes at given points of time
• MNX: Minimum and maximum of genes in the population

Summary statistic files follow this naming convention:

• AOI_MOI_QNT_qnt.csv

Where the main AOI was HLT (presence of mosquitoes) and the outputs (labels) are:

• TTI, TTO, WOP: Fraction's threshold for the metric to be true
• RAP: Fraction of present genotypes at given points (days) of the simulation
• POE: Probability of eliminating mosquitoes given the stochastic runs
• MNX: Min/Max and days at which these are achieved

Data Analysis Scripts

• ./PYF_preProcess.sh USR LND PLOTS_BOOL
• ./PYF_pstProcess.sh USR LND PLOTS_BOOL HEAT_BOOL
• ./PYF_clsPipeline.sh USR LND

Machine Learning Dashboard

After running the ML pipeline, load the dashboard with:

python PYF_clsApp.py PATH_TO_MODELS

Go to http://127.0.0.1:8050/

---