spinning planet

About IISER Kolkata

Detection and elimination of cattle-related pathogens are two of the main aspects of our project, Progenie, and that of another iGEM team, Namooste. The Indian Institute of Science Education and Research Kolkata (IISER-K)’s iGEM team, located in West Bengal, India, is working on project Namooste. Namooste’s goal is to detect and vaccinate against bovine mastitis, a deadly disease caused by Staphylococcus aureus that afflicts cattle. The SARS-CoV-2 pandemic has limited Namooste’s access to a wet lab, so Progenie supports the Namooste project by conducting experiments aimed at developing a real-time colorimetric testing kit consisting of a small RNA sensor for early detection and diagnosis of subclinical mastitis. In return, members of the IISER-K iGEM team shared their experience in mathematical modeling in both Python and Matlab’s SimBiology to help develop a computational model of how Progenie will behave in the environment.

Scroll down to learn more about our wet lab and dry lab collaborations!

We came to Team IISER-K with an initial design for an SEIR-based model.

The goal of this computational model was to provide insight into how Progenie could work and spread within a set population before we could get into testing each aspect. To model this spread, we started with an SEIR model and modified it to have the representative parts of our target population.

See ourmodel overviewfor more information on the current model.

See ourSEIR model overviewfor more information on SEIR models.

We initially recognized 4 distinct groupings of cells that would be relevant to our system:

  • 1. Those who could acquire φMINTO from initial phage infection (S_V)
  • 2. Those who were transduced with φMINTO and could then spread it to others (I)
  • 3. Those who could acquire φMINTO through bacterial conjugation (S_C)
  • 4. Those transconjugants who received φMINTO but don’t have the machinery to spread it to others (E).
  • Because the initial population would have only consisted of individuals who were susceptible to getting φMINTO through phage or conjugation, we had to devise a term in the model for the introduction of φMINTO-carrying phage. We first did this by making a compartment for phage that had influence on the flow of cells from S_V to I.

    Figure 1: The initial draft of the SEIR-like model for Progenie’s transmission dynamics.

    Team IISER-K’s modeling representatives helped us expand the initial design to be more realistic and informative.

    In our first modeling meeting with Team IISER-K, we walked them through this rough draft of a model. First, they asked questions that helped us identify the weak points of the design. Some initial pointers:

  • 1. The amount of phage present will not be constant - it will diminish over time as they bind and are destroyed.
  • 2. There probably won’t be much movement from the S_C group to the I group, if any. If the individual is a conjugative recipient, it won’t be a conjugative donor after receiving φMINTO.
  • 3. It would be easier and more informative to set the model in a closed system with a carrying capacity.
  • 4. The model will be more realistic if it includes aspects of Lotka-Volterra modeling for natural stochastic fluctuations in bacterial populations.
  • 5. Consider Kinetic Monte-Carlo or another stochastic model to simulate cell-to-cell conjugation interactions.
  • From these pointers, we made some initial modifications to each of their points respectively:

  • 1. Added arrows for the rates of phage binding and phage death, then used those rates to derive the equation for the rate of change of the amount of phage present.
  • 2. Removed the arrow and term C_I.
  • 3. Established a carrying capacity for a closed system.
  • 4. Together, we looked at simulations with and without Lotka-Volterra modeling to visualize the effect that it could have on our model. In trying to implement stochastic fluctuations to our model’s target population, we encountered difficulties and began to modify our model.
  • 5. The parameter set for simulating or model was very large and the range of values was exceedingly wide. For example, the conjugation rates were spanning over 11 orders of magnitude, creating a sensitivity issue for any stochastic model when it came to scaling for different systems.
  • 6. We decided to model Progenie using a series of deterministic ordinary differential equations and not have stochastic interactions.
  • Figure 2: Edits (red) made to the initial SEIR model plan during a Zoom meeting with Team IISER-K.

    After getting the model running, Team IISER-K helped us gain more insight from our model.

    After we built a model with the modified compartments and interactions using parameters that we found from literature, Team IISER-K recommended that we use our model to analyze a range of conjugation rates to test conjugation rates Progenie and determine what rates lead to success and failure. They recommended testing other values to satisfy our goal of understanding optimal parameters for Progenie model success.

    Figure 3: Team UCSC and Team IISER-K modeling teams meet over Zoom to discuss the model’s future directions.