Modeling and analyzing the kinetic of conjugative plasmids
- Prestes García, Antonio
- Alfonso Rodríguez-Patón Aradas Director
Defence university: Universidad Politécnica de Madrid
Fecha de defensa: 17 July 2017
- Fernando de la Cruz Chair
- Daniel Manrique Gamo Secretary
- María del Rocío López Igual Committee member
- Rafael Lahoz-Beltrá Committee member
- María Pilar Garcillán Barcia Committee member
Type: Thesis
Abstract
The bacterial conjugation, despite of the amount of molecular information available about its molecular aspects, does not have yet a well-established and structured systemic form which could allow a standard approach to the model building and the computational simulation task. This state of things implies that the modeling task are normally hard, repetitive, unreliable and extremely difficult to reproduce elsewhere by other research groups. Using a computational metaphor, we could state that bacterial conjugation is in a pre-structured programing era which makes the understanding of any particular implementation a very hard task and it also prevents the comparison of results from other models. This obviously hinders a leap forward in the use of the bacterial conjugation as a tool for doing computations with plasmid with plasmid encode mobile code. Even a very basic question, such as how should the conjugation process be represented in an individual-based model is still waiting for being answered. The problem also gets worst because the different backgrounds of synthetic biology research groups, which are interdisciplinary in their very nature, making hard to produce a homogenous view about the question mainly owing to the fact that the people from different domains use different vocabulary and have a different mental framework for understanding the process under study. The systemic view for any kind of process is much more than a simple accumulation of knowledge about system parts, which in the particular case of conjugation, is represented by the molecular and physiological aspects of bacterial life cycle. The action of generating an abstract representation modeling a system is what provides an holistic systemic view allowing to put focus on a precise point of the system but always taking into account the totality of the significant aspects. The modeling formalism itself for capturing the system properties also imposes limitations on the what can be observed or studied. In other words, the focus and the questions which can be answered are somehow dependent on the granularity of the modeling formalism. That is one of the greatest beauties of the individual-based formalism which forces the modeler to be concerned on the internal details of the entity being modeled as well as the time constraints and the order of the events. These factors provide much more information and allow to make conjectures about the internal strucure of individual bacterial cells when compared, for instance, to an ordinary differential equation model which can only expose the whole-population properies. In this thesis, we introduce an individual-based model for the conjugation process using, as the central point, the plausible link between the bacterial cell cycle and the time when the conjugative event is most likely to happen. The model was validated comparing the simulation outputs with the experimental data for different types of plasmids. Additionally, the model was assessed using the global sensitivity analysis methodology for providing a better understanding about the parameter uncertainty and the alternative model structures for representing the horizontal genetic transfer event. The thorough analysis seems to support the assumption that the conjugation dynamics is dominated by the parameter representing the time when the lateral gene transfer takes place. This conclusion is settled over two fundamental aspects. The first is related to the fact that, among the possible representations with respect to the temporal aspects, the best fit to the experimental data is achieved by the logic which takes into account the cell cycle. The second point is the sensitivity analysis which also indicates this parameter as significant. During the development of this thesis we had implemented the individual-based mode BactoSIM using the Repast framework. We had also implemented the GNU R packages EvoPER and R/Repast for the experimental setup and analysis. The tools developed for this work are available as opensource software. Moreover, a lateral contribution of this work is serving as a guide for other modelers in the systematic evaluation of their models.