MGSurvE: Mosquito Gene SurveillancE

A Python Package to Optimize the Surveillance of Genetically-Engineered Mosquitos Released in the Field

Main Operational Question

Given a heterogeneous environment and a limited number of traps, where should we place them?

MGSurvE's Approach

1) Workflow

2) Mosquito Components & Optimization Metric

3) São Tomé Demo

4) Status and Near Future

1) Workflow

2) Mosquito Movement

2) Optimization Metric

3) São Tomé Demo

4) Status & Near Future

São Tomé and Príncipe

Comparing Linked and Split Drive Constructs' Dynamics

Research Questions and Workflow

Massive panmictic experiments

Do sex-sorted releases make a difference?
Is there an optimal releases scheme?
How do the LDR and SDR time dynamics compare to each other?

Targeted spatial sims

What are the dynamics if we release in the south?
What are the dynamics if we release in the study sites?
How does LDR compare to SDR in spatial spread?

Panmictic Experiments

1) Releases Schemes and Fitness Costs

2) Feature Importance (sensitivity)

3) São Tomé Demo

4) Status and Near Future

1) Releases Schemes and Fitness Costs

2) Feature Importance (sensitivity)

Linked Drive

Split Drive

Spatial Experiments

1) Spread Dynamics

2) Example Video

3) "Take-Aways"

1) Releases Schemes and Fitness Costs

2) Example Video

3) "Take-Aways"

LDR: spreads across the island even if released in a confined region (until it's driven out by resistance).
LDR: needs less releases for effects in WOP.
SDR: provides good overall WOP if released across the island (provided the initial threshold is surpassed).
SDR: stays tightly confined if released in a specific region.
Both: Homing and resistant alleles fitness costs are the main gene-construct parameters determining the WOP.

Héctor M. Sánchez C.
Tomas León, Jared Bennett
Elijah Bartolome, Ayden Salazar, Lillian Weng, Xingli Yu, Joanna Yoo
John Marshall, David L. Smith
(sanchez.hmsc@berkeley.edu | chipdelmal.github.io)