Satya Arjunan/sandbox

From OpenWetWare

Jump to: navigation, search
Get PDF | Get log | Regenerate PDF | Export Source
Included files:

This is an email I wrote to Prof Gleb Oshanin, putting it here just for my
own records:

Dear Prof Oshanin,

Thank you for sending your paper about the determination of the reaction
probability p from the integral reaction rate, Q_N. I also greatly
appreciate your willingness to answer my questions. I apologize if I am
bothering you with trivial questions. I do not have any physics faculty
members here and I am doing my research independently. Most of my work is
related to biology. I hope to hear your opinion about a Monte-Carlo
simulation method that I have developed. I have attached the method in the
pdf file because it would be easier for you to view the equations.

With reference to the method, I would really be happy to hear your
comments about the following questions:

1. Is the method acceptable, i.e., correct, to reproduce both the reaction
and diffusion dynamics of particles in a liquid using a lattice?
2. If it is not correct, what do you think I should do?
3. Would it be possible for me to apply the lattice-based method described
in your paper to reproduce the reaction-diffusion dynamics of particles in
a liquid?
4. Do you know of any other lattice-based methods that you think is
5. Can I relate this method with the Collins and Kimball mean-field
approach since they say the local reaction rate at a point should be
equivalent to the diffusion of two particles into the point?

I really appreciate it even if you can give very short answers to these
questions. I am very good at programming (especially in C and C++
languages). If there is anything I can do to help you please let me know.

FYI, I am currently evaluating the survival probability of the target
particles in my method using the Equation 9 in the attached paper,
"Coarse-grained molecular simulation of diffusion and reaction kinetics in
a crowded virtual cytoplasm". I hope that it will be correct.

Best regards,
Satya Arjunan

Contents of the attached method in latex source:

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% User specified LaTeX commands.


Dear Prof Oshanin, my aim is to develop a lattice-based Monte-Carlo
simulation method that accurately reproduces the reaction-diffusion
dynamics of two mobile particles, $A$ and $B$, in a liquid with
a volume, $V\mathrm{m}^3$. The overall reaction in $V$ is given
by $A + B \xrightarrow{k_a} C$, where $k_a$ has the unit,
There are $N_A$, $N_B$ and $N_C$ number of $A$, $B$ and $C$
particles respectively. I am using a 3D lattice with a lattice spacing
of $a$, such that $a=R_A=R_B$, where $R_A$ and $R_B$ are the diameters
of the particle $A$ and $B$ respectively. To make all the particles
diffuse, i.e., randomly walk to a neighbor site in each time step
$\Delta t$, I let

       \Delta t=\frac{a^2}{6D}
\end{align}where $D$ is the diffusion coefficient of $A$ and $B$. When the
scavenger particle $A$ meets a target particle $B$ at its destination
site, they can react with a probability $p$. I determined $p$ from
the macroscopic rate constant $k_a$ with the following approach:
Say in the volume $V$ there are $N_S$ sites. At each simulation
step, the probability of finding a target particle $B$ by a particle
$A$ at its destination site is

\end{align}In a time step, the average number of target particles found by
scavenger particles is

       p_2&=N_A p_1\\
          &=\frac{N_A N_B}{N_S}
\end{align}Out of the $p2$ target particles found in a time step, if some
them react and $\Delta N_C$ particles are formed, then

       p&=\frac{\Delta N_C}{p2}\\
        &=\frac{\Delta N_C N_S}{N_A N_B} \label{p}
\end{align}In the liquid, we know from the law of mass action that

\end{align}In a very small $\Delta t$ within the volume V

       \Delta N_C = \frac{k_a N_A N_B}{V} \Delta t \label{deltaNc}
\end{align}Substituting \eqref{deltaNc} in \eqref{p}

       p&=\frac{k_a N_S}{V}\Delta t

Personal tools