| The cell is a basic unit of life, and it is also the key to open the mystery of life, conquer diseases and benefit the human health which makes it essential to research cell intensively. A cell contains a lot of biological macromolecules such as protein, nucleic acid and so on. A series of complicated and precise biochemical reactions among these macromolecules form the cell macromolecular interaction network. In order to monitor and study this interaction network directly and simultaneously, theoretical biologists have built different virtual cell models, some models have disadvantages of inaccurate calculation result, some can correctly describe the intracellular movement, but they have disadvantages of extremely expensive computation which make these models not be applied to a higher level of research on cells.Based on the former researches, this thesis introduces electronic and van der Waals interactions, and develops coarse-grained cytoplasm models and cell models which include206kinds of essential proteins of E.coli. By analyzing the diffusion movement of our validation model, it shows that our models are less time-consuming, and the results are reliable, which indicates that our models can correctly describe the intracellular movement, and provide a theoretical basis and technical support for further research of macromolecules'movement in cell.The main contents and results of this thesis are as follows: 1. We constructed a 'minimal proteome' coarse-grained cytoplasm model which contains206E.coli essential protein types. we analyzed this model and found that both proteins type and protein copy number of206proteins are in the bimodal distribution under different pI value. There is a negative correlation between the protein net charge at pH7and the molecular weight. That's to say most of the positively charged proteins are small, while large proteins tend to have large negative charges.2. The11cytoplasm models that we constructed were simulated based Langevin equation, and all the dynamics simulations were performed by NAMD. We used CHARMM Mathematica and Python to analyze the results of the simulations, and found that most proteins are involved in a complicated protein-protein interaction network within cell.3. We researched the diffusion movements and did collision analysis of the normal concentration of proteome models NM1-3, M8and M16to further analyze the protein-protein interaction network. The results showed that the mass and charge of proteins affect their diffusion properties:the diffusion constants for charged particles increase linearly with the ratio of charge to mass. Proteins interact with each other by collision, and the protein collision rate increases linearly with the absolute charge of the proteins. As a result, proteins with near neutral charge have the lowest collision rates.4. Two molecules were considered to be in a cluster and in contact with each other if the distance between them is equal or less than12A, thus all molecules in one cluster were connected with each other directly or indirectly. We compared the clusters of the six random control sets with the minimal proteome model NM1-3, and found that the proteins in the minimal proteome model form larger clusters than the random control sets, and the size of the clusters fluctuates slightly over time, indicating that the cluster is dynamic.5. Based on the cytoplasm models, we added RNA and circular DNA molecules to our models to construct the E. coli virtual cell. Using the same simulation and analysis method to study the cell model, we found that the protein concentration and charge also affect proteins movement and interaction inside the cell model. Thus, these conclusions can be applied to the whole cell.Our study also suggests that'proper'populations of negatively and positively charged proteins are important for maintaining a protein-protein interaction network in a cell. And the charge and mass fo proteins can contribute to pre-organize the protein-protein interaction network. |
PDF Full Text Request