| 1 |
tim |
2685 |
\chapter{\label{chapt:conclusion}CONCLUSION} |
| 2 |
tim |
2877 |
|
| 3 |
|
|
This dissertation has presented the work and research accomplished |
| 4 |
|
|
over the course of the past five years. Chapt.~\ref{chapt:RSA} |
| 5 |
|
|
illustrated the application of statistical mechanical principles to |
| 6 |
|
|
the solution of an experimental problem. Given a relatively simple |
| 7 |
|
|
model, could the differences in experimental coverage be accounted |
| 8 |
|
|
for? By utilizing Monte Carlo modeling methods, I wrote several |
| 9 |
|
|
simulations that provided a clear explanation for the majority of |
| 10 |
|
|
the experimental evidence. Namely, that overlapping, or |
| 11 |
|
|
interdigitating molecules were responsible for the high surface |
| 12 |
|
|
coverage of the ``umbrella'' silicon phthalocyanines, while the |
| 13 |
|
|
``octopus'' molecules, which do not overlap, have a much lower |
| 14 |
|
|
coverage. |
| 15 |
|
|
|
| 16 |
|
|
Chapt.~\ref{chapt:oopse} presented the work necessary to the |
| 17 |
|
|
solution of more complex research problems. It became clear, after |
| 18 |
|
|
the work on the RSA simulations, that even simple simulations can |
| 19 |
|
|
take a great deal of programming. If I were to simulate the bilayer |
| 20 |
|
|
systems using a coarse-grained model, a sophisticated molecular |
| 21 |
|
|
modeling program was required. After examining the options available |
| 22 |
|
|
at the time, writing an advanced simulation program was the only |
| 23 |
|
|
realistic solution. |
| 24 |
|
|
|
| 25 |
|
|
Working with my lab-mates, I have developed our simulation program |
| 26 |
|
|
({\sc oopse}) to the point where it is capable carrying out many |
| 27 |
|
|
unique simulations. It properly integrates rigid body dynamics. It |
| 28 |
|
|
has the capability to model dipoles rather than point charges only; |
| 29 |
|
|
and is therefore able to integrate larger systems for longer times |
| 30 |
|
|
than any other current modeling package. It can use the MPI |
| 31 |
|
|
interface to calculate the long range forces across multiple |
| 32 |
|
|
processors. And lastly, it is open source, so that other scientists |
| 33 |
|
|
may use the code and contribute back to the project and enhance the |
| 34 |
|
|
program. |
| 35 |
|
|
|
| 36 |
|
|
Lastly Chapt.~\ref{chapt:lipid} presented the research that became |
| 37 |
|
|
the main design goal of {\sc oopse}. The lipid model was under |
| 38 |
|
|
continual development while {\sc oopse} was being written. In fact |
| 39 |
|
|
many of the integration ensembles in {\sc oopse} were specifically |
| 40 |
|
|
written to continue development of the lipid model. The set of |
| 41 |
|
|
simulations presented in the chapter have shown that the model is |
| 42 |
|
|
capable of spontaneously forming bilayers. It also explored the |
| 43 |
|
|
properties of the bilayers over a range of temperatures. Further |
| 44 |
|
|
extensions to the model will likely include exploration of how head |
| 45 |
|
|
group dipole strength and varying chain configurations (one chain |
| 46 |
|
|
versus two chains, etc.) affect the structure and dynamics of the |
| 47 |
|
|
lipid bilayer. However, at this point there are many questions that |
| 48 |
|
|
can now be explored with this model such as how do small molecules |
| 49 |
|
|
include themselves, and diffuse through the bilayer? Or what are the |
| 50 |
|
|
characteristic events that lead to the formation of special lipid |
| 51 |
|
|
phases (i.e.~the ripple phase)? It is my hope that my contributions |
| 52 |
|
|
to this research will make it possible to answer these important |
| 53 |
|
|
questions and many others. |
