BE.011/2.

772J
Statistical Thermodynamics of Biomolecular Systems
Spring 2004
Griffith/Hamad-Schifferli
Problem Set #1
Due: 2/11/04

7.) You have 4 boxes and 3 different (i.e., distinguishable) balls, which are Red, Yellow, and
Blue.
a) If there is no restriction on the number of balls per box, how many different ways can the balls

be put in the boxes?

b) If we impose a restriction that the first box must contain 2 balls, and the second box must

contain 1 ball, how many ways can this be achieved? First, write out all the possibilities. Then,

write out the general formula you would use to calculate it, and see if it matches.

c) What is the probability that the situation in b) (2 balls in box 1, 1 in box 2) was achieved if the

balls were thrown in randomly, as in a)?

8.) You have a genome that is exactly 1 ? 109 bases long. You would like to choose a DNA
sequence out of the genome that is unique. What is the minimum length of the sequence such
that it could be unique (i.e., it is possible that it does not occur anywhere else in the genome)?

BE.011/2.772J
Statistical Thermodynamics of Biomolecular Systems
Spring 2004
Griffith/Hamad-Schifferli
Problem Set #2
Due: 2/11/04

Dill 2.1, 2.2, 2.3, 2.4, 2.5, 2.6

BE.011/2.772
Problem Set 3
Due March 3, 2004
1. Dialysis is common laboratory procedure used to remove low molecular weight solutes from protein
solutions. Example situations include: removal of removal of free, unreacted label after
fluorescently labeling a protein; and removal of urea from a denatured protein solution to allow
refolding of the protein. The protein solution is sealed inside a dialysis bag or tubing and placed in
container with a large excess of pure water, as shown in the figure. The system is then allowed to
come to equilibrium. For this problem, consider an idealized system where the only solute in the
dialysis bag is the low molecular weight solute (i.e., no protein is actually present.)
(a) Explain why the combined system of the protein solution (system A) and the water (system
B) may be considered to be an isolated system.
(b) Use a lattice model, with each individual water molecule occupying a single site in the
lattice, to show that the entropy of System B is zero at the start of the experiment. Designate
the number of water molecules NWB and the number of total lattice sites MB.
(c) Use a lattice model to express the entropy

of system A at the start of the experiment

in terms of the number of water molecules

in system A, NwA, the number of solute

particles, Ns, and the total number of lattice

sites in system A, MA. For your final

expression, use an appropriate

approximation to eliminate factorial

expressions. You may assume that the

water molecules and solute molecules are

comparable in size.

(d) Now write an expression for the entropy of

the combined system under the conditions where water and solute can freely pass through the
dialysis membrane. For every solute molecule that crosses over to system B, a water
molecule must cross over to system A. (You may want to keep track of the number of solute
molecules by noting that NS =NSA+ NSB and the number of water molecules by noting that
NWA + NWB = NW).
(e) Show that you can write the entropy expression in terms of one system variable (i.e., one
parameter that changes value during the progress of the experiment), and briefly describe
how to determine the equilibrium condition in terms of that variable and constants in the
system. [You do not need to work out the equilibrium condition (lots of algebra involved)
but feel free to predict what it is if you like.]
(f) Is work done in the approach to equilibrium? Explain why or why not.
(g) Does the internal energy U of the system change? Explain why or why not.

BE.011/2.772

Problem Set 4

Due March 8, 2004

Dill 8.3, 8.4, 8.5, 8.6

BE.011/2.772

Problem Set 5

Due March 17, 2004

Dill 9.1, 9.3, 9.8 and 9.10

Continued on next page....

 U
U
We are given that
= 0 , and we are asked to show that this implies that CV =
is

V T
T V
CV
U

not a function of V. For this, we need only prove that

=
= 0 . But we know that we can

V
T V
V
CV

U
U

(0) = 0 . Thus,
=
permute the order of the derivatives, so we get that

=
V
V T V T V T T
the constant volume heat capacity of an ideal gas is independent of volume.

9.3

f
9.8 The first thing to note is that if we take the cross derivatives of dU, we end up getting

p0 E ,S
p
rather than 0 which we want. We need to first transform our equation to a new energy variable
f E ,T
X(T,f,E).

dU = TdS +
fdl + Edp0
X
= U TS fl Ep0
dX = SdT
ldf p0 dE
p
Now if we take the cross derivatives we can get an expression for 0 .
f E ,T

p0
l

=

f
E ,T E
f ,T

This is useful, because we are likely able measure how the length of the crystal changes with the field
l

.
E f ,T

9.10
a) The first thing to note in this question is that we are considering T and p as variables (which we end
up holding constant). That means that we need to use the Gibbs free energy.

G
= H TS =
U + pV TS
dG = dU + pdV +
Vdp TdS SdT
= SdT +
Vdp + fdl
Now we want to create a Maxwell relation between S and f in order to get a function for S.

 2G 2G

lT Tl
G
G

l T l , p T l T , p
f
S
=

l T , p T p ,l

S
(aT (l l0 ))
=
T
l T , p
= a (l l0 )
l

S (l ) = a(l l0 )dl
l0

a (l l0 )
2
So the entropy is proportional to the square of the extension distance.
S (l ) =

H = U + pV
dH = dU + pdV + Vdp
= TdS + Vdp + fdl
dH
dS

= T + f
dl T , p
dl T , p

= T ( a(l l0 )) + aT (l l0 )
=0

So H is constant for all extensions l.

b) Since we are told that it is an adiabatic process (q=0, so dU=w=fdl), we know that we need to start
with the expression for internal energy U. We also know that by definition, for constant volume
dU=CvdT.
dU = fdl

Cv dT = aT (l l0 )dl
Cv
dT = a (l l0 )dl
T
T2

C
T Tv dT = l a(l l0 )dl
1
0

T2 a (l l0 )2
Cv ln =
2
T1
a(l l0 )2
T2
= exp

T1
2Cv

Problem Set 6

2.772/BE.011

Dill & Bromberg: 10.1, 10.3, 10.6, 10.11

Problem Set 7
2.772/BE.011

Dill & Bromberg: 10.7, 10.8, 11.2, 11.5, 11.11(a)

BE.011/2.772 Homework Week of April 12, 2004

2. Integrins are a family of transmembrane receptors that mediate adhesion between cells and
extracellular matrix by binding to specific ligands in extracellular matrix molecules, and then linking to
the cytoskeleton inside the cell to form a continuous physical bridge. When blood platelets are
activated, the integrin aIIb3 on the surface of the platelets binds to peptide sequences containing the
sequence RGD (Arg-Gly-Asp) in the molecule fibrinogen. Upon binding to ligand, the integrins are
driven to cluster together, facilitating strong adhesion and degranulation of platelets, resulting in clot
formation. Clot formation in blood vessels can have severe pathophysiological consequences, including
stroke and embolism, and thus drugs that inhibit integrin-mediated functions on platelets are now being
developed. One drug, Integrilin, is a small soluble peptide that binds to the integrin and blocks binding
to fibronectin.
Hantgan et al1 sought to determine whether enthalpy or entropy drives the formation of integrin
aIIb3 clusters after ligand binding. They used 3 different RGD-containing ligands (including the drug
Integrilin) to form receptor-ligand complexes, and measured dimerization of ligand-bound receptors as a
function of temperature in the range 20-40C using a variety of methods. The 3 ligands are cHArGD,
cRGD, and Integrilin. The dimerization reaction of ligand-bound receptors can be written as
2RL dimer
The values of the equilibrium constant K as a function of temperature for each of the 3 ligands are given
in the following table
T C

Ka,cHArGD (M-1)

Ka,cRGD (M-1)

Ka,Integrilin (M-1)

20
25
30

1.79 x 104
5.9 x 104
1.95 x 105

1.79 x 104
7.08 x 104
2.73 x 105

5.0 x 104
8.6 x 104
1.47 x 105

35
40

5.97 x 105
1.79 x 106

9.85 x 105
3.47 x 106

2.45 x 105
4.00 x 105

(a) Which ligand leads to the greatest degree of dimerization of the integrin-ligand complex at 20C?
At 40C? Calculate the fraction of total integrin-ligand complexes that are dimerized at
equilibrium at 20C and 40C if the starting concentration of integrin-ligand complexes is 1 x 10-6
M.
(b) For each of the ligands, calculate the Gibbs free energy change at 20C.
1

Hantgan, R.R., Lyles, D.S., Mallett, T. Conn, Rocco, M., Nagaswami, C., and Weisel, J.W. , J. Biol.
Chem., 278(5), 3417-3426 (2003).

(c) The values of H and S are presumed independent of temperature over this range. Using the
data in the table, calculate H and S values for each of the ligands.
(d) Is the dimerization driven by entropic effects or enthalpic effects? Provide an analysis of the
driving force from a molecular perspective.

Problem Set 9

Dill & Bromberg, chapter 16, problems 2, 7, 9, 12, 14, 16

Continued on next page....

vesicle. The osmotic flow could be reversed by a sufficient concentration of a different

nonexchangeable protein species B outside the membrane. The osmotic flow could be

reversed because at low concentrations each distinguishable species of nonexchangeable

component reduces the chemical potential independently of every other species.

16.14) Osmosis in plants. Plants must lift water from ground level to their leaves.

Consider the roots as a capillary with radius 0.01cm whose walls are a semipermeable

membrane with pure water on the outside and a solution inside with solute mole fraction

x = 0.001. The solution inside has density 1 g cm-3. What is the height of the solution at

room temperature?

For the ideal solution given, the osmotic pressure across a membrane with chemical

potential difference is given by:

= where is the molar volume of the solute. (equation 16.31 in text)

And, from the equation on the bottom of page 292 (derived from equation 16.39), and

from the free energy of transfer equation, 16.41, we know that:

= kTxsolute

We need it in this form because we are given xsolute = 0.001.

Remember that is equivalent to energy, where the total energy = kinetic energy +

potential energy. (This is just one of the many equations we know for energy, but this

equation is relevant here because we need to know about the downward force on the

water moving up.) At equilibrium, the water is no longer rising, so there is no kinetic

energy, so the free energy equals the gravitational potential energy at equilibrium:

= mgh.

We do not have the mass, but we can get mass from the volume and the density:

= Vgh

The volume of the cylinder is given by: V=r2h so,

= gr2h2.

Now we know all the parameters in these two equations so we can set them equal to

each other:

gr 2 h 2 = kTxsolute
h2 =

kTxsolute
gr 2

g cm3
)(0.001)
2
s
=
g
cm
(1 3 )(980 2 )(3.14)(10 4 cm 2 )
cm
s
(4.14 x1014

h 108 cm
Therefore, although osmotic pressure does contribute to water transport inside of trees,
other factors, such as transpiration of the water from the leaves, are much more
important.
16.16) Ethane association in water. For the association of ethane molecules in water,
h0 = 2.5 kcal mol-1 at T=25C. We want to know how the association constant K
changes with temperature.

1
h 0 Ts 0 , we would be able to get a value
RT
for K if we knew s. Though we do not know s, we can still get an expression for the
change of K with temperature (as long as we assume that h0 and s0 do not depend on
T.
h0 s 0
+
ln K =
RT
R
0
2500cal / mol
ln K h
=
=
= 0.0142 K 1
2
RT
(1.987cal / molK )(298 K ) 2
T
As T increases, K increases, so the hydrophobic effect gets stronger.

From the vant Hoff equation, ln K =

Page 1 of 14
Name:

BE.011/2.993J
Spring 2003
FINAL EXAM
May 19, 2003
You have 3 hours for this exam.
CLOSED BOOK
4 pages of notes allowed
1. (60 points)
2. (20 points)
3. (10 points)
4. (40 points)
5. (65 points)
6. (30 points)
total (225 points)

Some constants you may need:

k = 1.38 ? 10-23 J K-1
R = 8.314 J K-1 mol-1
h = 6.626 ? 10-34 J s
1atomic mass unit (amu) = 1.66 ? 10-27 kg

Page 2 of 14
Name:

L
?

L
?

1.) (60 points total) The protein below has four binding sites (? , ? , ?, ? ) for the ligand L (look
familiar?). Assume that the association and dissociation constants are equal.

It has energies associated with each macrostate L, where L is the # of ligands:

#L
0
1
2
3
4

energy
8?
4?
2?
1?
0

1a.) (10 points) Plot an energy level diagram, including all microstates. Label values on the y
axis in terms of the energy unit ?.

Continued on next page....

Page 3 of 14
Name:

1b.) (10 points) Write down an expression for the partition function q. Calculate its value at a
temperature where kT = 3?.

1c.) (5 points) What is the highest macrostate that is populated at this temperature kT = 3?? How
many of the microstates in this macrostate are populated?

1d.) (10 points) Calculate pL for L = 0,1,2,3,4 at a temperature where kT = 3?. What is ? LpL?

Page 4 of 14
Name:
1e.) (10 points) Calculate <E> at a temperature where kT = 3?, (you can leave it in terms of ?).

1f.) (5 points) What is the entropy, S, at a temperature where kT = 3??

1g.) (5 points) What is q as T? ? ?

Page 5 of 14
Name:
1h.) (5 points) What is S as T? ? ?

2.)(20 points total) Heat capacity of a molecule

You have the molecule glucose, which has 66 degrees of vibrational freedom. Plot its heat
capacity as a function of temperature, making sure to denote on the plot

values for heat capacity (in units of R) on the y axis

where the rotational and vibrational temperatures, ? rot and ? vib, are on the x-axis.

Cv/R

Page 6 of 14
Name:

3.) (10 points total) Ribonuclease A is a protein that has 124 amino acids.
3a.) (5 points) What is the probability that this sequence is unique?

3b.) (5 points) Assume that residues 1-20 are conserved (exactly the same). How many different
homologies are possible for RNase A?

Page 7 of 14
Name:

4.) (40 points total) Equilibrium between two states. The reaction
A? ? B
has the equilibrium constant K. The energy levels for A and B are shown below:
E

8?0
6?0

6?0

5?0
4?0

4?0
3?0

2?0
0

2?0
? =2?0

A has a fixed energy level spacing of 2? 0 , with a ground state energy of 0. B has a fixed energy
spacing of ? 0 with a ground state energy of 2? 0 .
4a.) (10 points) Calculate K for the reaction, assuming kT=2? 0 .

Page 8 of 14
Name:
4b.) (10 points) What is K as T? ? ?

4c.) (5 points) What side of the reaction is favored as T? 0?

4d.) (10 points) Now suppose the ground state energy of B was also 0 (i.e., ? =0). Calculate K
using kT=2? 0 gain.

Page 9 of 14
Name:
4e.) (5 points) Lets pretend A? ? B is an isomerization reaction, i.e., a molecule changing only
in shape such as Helix ? ? Coil, and that the energy levels are vibrational. Which form of the
molecule has stiffer bonds (higher bond spring constant, k)?

Continued on next page....

Page 10 of 14
Name:

5.) (65 points total) Were going to test the effects of CO gas on cells. Our experiment utilizes a
thin, two-dimensional gas of the diatomic molecule CO in a petri dish that has dimensions a ? b.
5a.) (5 points) Write down an expression for the full partition function for the CO molecule in
terms of constants (include all degrees of freedom).

5b.) (10 points) Calculate a value for the qtrans using T = 37C, a = b = 3cm. Use the mass values
of mO = 16 g/mol and mC = 12 g/mol.

5c.) (10 points) What is the U per molecule (U/N)?

Page 11 of 14
Name:

5d.) (10 points) Now lets compare the rotational properties of CO with that of a DNA oligo.

Experimentally you find that the first four energy levels for the diatomic CO to be the following
energies:
0J
7.64 ? 10-23 J
2.29 ? 10-22 J
4.58 ? 10-22 J
Determine the bond length for the molecule.

5e.) (10 points) Calculate a value for qrot using T = 37?C. What is the rotational temperature for
CO?

Page 12 of 14
Name:

5f.) (5 points) If we were make the same measurements in e) at a different temperature, what
would happen to the values of the energies observed?)

5g.) (10 points) Now lets look at the DNA molecule. The oligo is 12 nucleotides long, and each
strand has a molecular weight of 3832 g/mol. Calculate the energy levels of the first four
rotational states for the DNA oligo. Assume the length is 3.4 per base pair. Also assume it is a
rigid rod with rotation around the center as shown in the figure, so that I = 1/12 ML2 , where L is
the entire length of the molecule.

5h.) (5 points) Calculate the rotational temperature, ? rot , for this rotating DNA oligo. How does
this compare to the ? rot for CO?

Page 13 of 14
Name:

6.) (30 points total) Solutions and Mixtures.

6a.) (10 points) Pure liquid hexane is suspended in a sealed container in the absence of gravity
(ok, slightly fictitious! This situation does model a real situation mentioned at the end). The
temperature is constant. The hexane does not touch any of the surfaces of the container and is
literally suspended in space. Think back to our discussion of surface tension and the free energy
associated with creating surfaces of condensed phases. At equilibrium, what geometry (i.e.,
shape) is the liquid hexane and why?

6b.) (10 points) The hexane in part a is mixed with water and the mixture is suspended in the
sealed container in a gravity-free environment, just as in part a. As you probably already
intuitively know, wA-A >>> wBB > wAB where A refers to water and B to hexane, and w refers to
the bond energy (in fact, a negative number. We are following the convention of the text here to
indicate that the A-A bonds are stronger, i.e., a bigger negative number). Sketch the equilibrium
configuration of the system, presuming the liquids do not contact any of the container surfaces
and are literally suspended in space. Where is the water relative to the hexane, and what
geometry (shape) does the liquid have? You may assume that there are comparable total volumes
of water and hexane present.

Page 14 of 14
Name:

6c.) (10 points) You were not told anything about the relative volume of liquid to the total
volume. If the container is sealed after introduction of the liquids, and held at constant
temperature, are there any restrictions on how much liquid is added initially so that liquid is
present when equilibrium is reached?

Relation to real problems: Malcolm Steinberg, a developmental biologist at Princeton

University, proposed that some forces controlling development of multicelluar organisms are
analogous to those described in part b. He in fact showed that when two cell types that express
the same cell-cell adhesion receptor are mixed in suspension, they behave pretty much the same
as the liquids described in part b if cell type A has a lot more of the receptors than cell type B.

Thats it! Have a great summer!!!

BE.011/2.772
Quiz 2
March 20, 2003
1. Dialysis is common laboratory procedure used to remove low molecular weight solutes from protein
solutions. Example situations include: removal of removal of free, unreacted label after
fluorescently labeling a protein; and removal of urea from a denatured protein solution to allow
refolding of the protein. The protein solution is sealed inside a dialysis bag or tubing and placed in
container with a large excess of pure water, as shown in the figure. The system is then allowed to
come to equilibrium. For this problem, consider an idealized system where the only solute in the
dialysis bag is the low molecular weight solute (i.e., no protein is actually present.)
(a) Explain why the combined system of the protein solution (system A) and the water (system
B) may be considered to be an isolated system.
(b) Use a lattice model, with each individual water molecule occupying a single site in the
lattice, to show that the entropy of System B is zero at the start of the experiment. Designate
the number of water molecules NWB and the number of total lattice sites MB.
(c) Use a lattice model to express the entropy of system A at the start of the experiment in terms
of the number of water molecules in system A, NwA, the number of solute particles, Ns, and
the total number of lattice sites in system A, MA. For your final expression, use an
appropriate approximation to eliminate factorial expressions. You may assume that the water
molecules and solute molecules are comparable in size.
(d) Now write an expression for the entropy of the combined system under the conditions where
water and solute can freely pass through the dialysis membrane. For every solute molecule
that crosses over to system B, a water molecule must cross over to system A. (You may want
to keep track of the number of solute molecules by noting that NS =NSA+ NSB and the number
of water molecules by noting that NWA + NWB = NW).
(e) Show that you can write the entropy expression in terms of one system variable (i.e., one
parameter that changes value during the progress of the experiment), and briefly describe
how to determine the equilibrium condition in terms of that variable and constants in the
system. [You do not need to work out the equilibrium condition (lots of algebra involved)
but feel free to predict what it is if you like.]
(f) Is work done in the approach to equilibrium? Explain why or why not.
(g) Does the internal energy U of the system change? Explain why or why not.

Initial state

Equilibrium
state

2. In class we considered the text example of a polymer comprising 4 monomer units that can exist in
a compact state or an open state. The compact state arises from an attractive bond between the
monomers at the ends of the chain (see figure). This bond is characterized by an energy . If this
bond has an energy of 588 x 10-23 J, what is the free energy of the compact system at 310K? What is
the free energy of the open system at 310K? Will there be more open or compact chains at 310K?
The value of k is 1.38 x 10-23 J/K.
3. One gram of water is heated in a calorimeter from 274K to 314K. The following data are obtained:
Temperature

Cumulative Heat added (bomb

calorimeter)

Total Heat added (constant

pressure calorimiter)

274
294
314

0 cal
19.7 cal
39.5 cal

0 cal
20 cal
40 cal

What is the change in entropy when the temperature of 1 gm of water is raised from 274 to 314K at
constant volume? At constant pressure?

Page 1 of 4
Name:

BE.011/2.993J
Spring 2003
QUIZ I
February 26, 2003
You have 1 hour for this exam.
CLOSED BOOK
1 page notes allowed

1 (20 points)
2 (50 points)
3 (30 points)
total (100 points)

1.) (20 points) DNA sequence

...GTAGCCGTGAATCGATGGTCCAAATACCG...

|------->?

You have a genome that is exactly 1 109 bases long. You would like to choose a DNA
sequence out of the genome that is unique. What is the minimum length of the sequence such
that it could be unique (i.e., it is possible that it does not occur anywhere else in the genome)?

Page 2 of 4
Name:

b
+

L
a

a) (15 points) Calculate W and the entropy (in units of k) for the situation in which
i) 0 ligands are bound (NL = 0)

ii) 1 ligand is bound (NL = 1)

iii) 2 ligands are bound (NL = 2)

iv) 3 ligands are bound (NL = 3)

v) 4 ligands are bound (N L = 4)

2.) The protein below has four binding sites (a, b, g, d) for the ligand L. We would like to find its
equilibrium binding population. For now assume that the association and dissociation constants
are equal.

Page 3 of 4
Name:

b) (5points) Which states have the highest entropy?

c) (5 points) Which states have the lowest entropy?

d) (5 points) Which state will you find it in at equilibrium?

e) (20 points) Lets say that the binding constants are not equal, i.e., ligand L has a higher
probability of being bound:
pbound = 0.75
punbound = 0.25
Plot the probability distribution for all of the states, p(i).

Continued on next page....

Page 4 of 4
Name:

3.) DNA has 3 different possible configurations: A, B, and Z:

Let the scores for each configuration e i be : e A = 1 , e B = 2, e Z = 3.

a) (10 points) Write down the partition function, q. (you may substitute x = e-b )

b) (10 points) Write down the expressions for pA and pB and pZ.

c) (10 points) What is the variance of the distribution, <s2 >, if all three configurations are
equally possible?

Page 1 of 6
Name:
BE.011/2772J
Spring 2004
QUIZ I
February 23, 2004
You have 1 hour for this exam.
CLOSED BOOK
1 page notes allowed

1 (10points)
2 (5 points)
3 (25 points)
4 (60 points)
total (100 points)

k = 1.38 10-23 J/K

Continued on next page....

Page 2 of 6
Name:
1.) ( 10 points) ELVIS is everywhere
a.) Given 20 naturally occurring amino acids, what is the probability that the amino acid
sequence ELVIS occurs in a stretch of a protein sequence?

b.) What is the probability if the order of the amino acids did not matter, i.e., VLSEI, etc.?

2.) (5 points) Protein folding.

a.) A protein is a linear chain of amino acids. The amino acid has two torsional angles than can

vary around the carbon, and . Due to sterics, and have 3 possible configurations apiece,

yielding 3 3 = 9 possible configurations per amino acid. Pictured below is a three residue

peptide:

N
H

OH

N
R

If a protein has n =100 amino acids, how many different configurations are possible?

Page 3 of 6
Name:
3.) (25 points) Conditional probabilities of the genetic code
a) A codon is a sequence of 3 nucleotides that specifies a particular amino acid. How many
codons are possible out of the 4 nucleotides?

b) Using the table for the genetic code below, what is the joint probability of obtaining a G in the
second position (G2) and A in the first (A1)?
Second Letter

First
Letter
U

UUU

Phe

UCU

Ser

UAU

Tyr

UGU

Cys

UUC

Phe

UCC

Ser

UAC

Tyr

UGC

Cys

UUA

Leu

UCA

Ser

UAA

Stop

UGA

Stop

UUG

Leu

UCG

Ser

UAG

Stop

UGG

Trp

CUU

Leu

CCU

Pro

CAU

His

CGU

Arg

CUC

Leu

CCC

Pro

CAC

His

CGC

Arg

CUA

Leu

CCA

Pro

CAA

Gln

CGA

Arg

CUG

Leu

CCG

Pro

CAG

Gln

CGG

Arg

AUU

Ile

ACU

Thr

AAU

Asn

AGU

Ser

AUC

Ile

ACC

Thr

AAC

Asn

AGC

Ser

AUA

Ile

ACA

Thr

AAA

Lys

AGA

Arg

AUG

Met

ACG

Thr

AAG

Lys

AGG

Arg

GUU

Val

GCU

Ala

GAU

Asp

GGU

Gly

GUC

Val

GCC

Ala

GAC

Asp

GGC

Gly

GUA

Val

GCA

Ala

GAA

Glu

GGA

Gly

GUG

Val

GCG

Ala

GAG

Glu

GGG

Gly

c) Calculate the degree of correlation for obtaining a Serine given G2. Is it negatively/
positively/not correlated, or mutually exclusive?

Third
Letter
U
C
A
G
U
C
A
G
U
C
A
G
U
C
A
G

Page 4 of 6
Name:
4.) (60 points) Protein ligand binding
A protein has M = 3 sites for binding a ligand. The sites are indistinguishable from each other, as
are the ligands.
NL=0

NL=1

NL=3

NL=2

a) Express the number of ways NL ligands can be arranged in M sites, W (M, NL).

b)Calculate the multiplicity and also the entropy for the following states. NL= 1 means that one
ligand is bound, NL=0 means no ligands are bound, etc. You may leave the entropy in terms of
the Boltzmann constant, k.
(i) NL =0

(ii) NL = 1

(iii) NL = 2

(iv) NL = 3

Page 5 of 6
Name:
c) Plot the probability distribution as a function of the number of ligands, p(NL).

d) Calculate the variance of the distribution.

Page 6 of 6
Name:
e) Upon addition of a cross linker, the protein can dimerize to form the following dimer (pictured
below). The dimer can bind up to 4 ligands.
NL=1

NL=2

NL=1

cross linker

What is the entropy of the dimeric protein where NL =2 (pictured)? You may leave in terms of k.

f) Compare the entropy of two proteins as monomers (left) with NL=1 with the entropy of the
dimer (right). What is the change in entropy, S, going from monomeric to dimeric, in terms of
k? Based on your calculation, is the system more likely to be in the monomeric or dimeric form?

Quiz Review Problems: 3.2, 6.1, 6.4, 6.10 Solutions

3.2 Equalizing energies. For the two 10-particle two-state systems of Example 3.4,
suppose the total energy to be shared between the two objects is U = UA + UB = 4. What
is the distribution of energies that gives the highest multiplicity?
For this problem, dont make any assumptions about the starting positions of the particles
(as they have in example 3.4), you just want to know how to distribute the particles in the
two systems so that the total U=4. However, we can assume the systems are of the same
format (the first energy level is e=0 and the second is e=1) and that there are again 10
particles in each system. And, lets denote the number of particles in state i by ni. So, the
multiplicity of each system individually is:

10
!
so the total multiplicity is:
W
=

n
n
!(10

)!
1

10!
10
!
(10!)
2

Wtot =

n1, A !(10 n1, A )! n1, B !(10 n1, B )!

n1, A !(10 n1, A )!(4 n1, A )!(6 + n1, A )!
We can evaluate this for all possible values of n1,A:
n1,A
W
0
210
1
1200
2
2025
3
1200
4
210

6.1 Calculating the entropy of dipoles in a field. You have a solutions of dipolar
molecules with a positive charge at the head and a negative charge at the tail. When there
is no electris field applied to the solution, the dipoles point north (n), east (e), west (w), or
south (s) with equal probabilities. The probability distribution is hown in Figure 6.7(a).
However when you apply a field to the solution, you now observe a different distribution:
the dipole points north with probability 7/16, east or west with probability and south
with probability 1/16.
(a) What is the polarity of the applied field? (In which direction does the field have
its most positive pole?)
The positive heads of the dipole should point away from the positive pole of the
field. Therefore the field must have its positive pole to the south.
(b) Calculate the entropy of the system in the absence of the field.

Use the formula: S = k pi log p

i
i

Since all directions are equally likely, S = -k*4*( log ) = klog4.

(c) Calculate the entropy of the system in the applied field.
7
1
1
1
1
7
7
S = k log + 2 * ( log ) + log = k(3 log 2 log 7) 1
.228k
16
4
4 16
16

16

16
(d) Does the system become more ordered or disordered when the field is applied?

The entropy decreases when the field is turned on so the system becomes more
ordered.
6.4 Calculating the entropy of mixing. Consider a lattice with N sites and n green
particles. Consider another lattice, adjacent to the first, with M sites and m red particles.
Assume that the green and red particles cannot switch lattices. This is state A.
(a) What is the total number of configurations WA of the system in state A?

N!
M !

W A = W N * WM =

n!( N n)! m!(

M m)!
(b) Now assume that all N+M sites are available to all the green and red particles.
The particles remain distinguishable by their color. This is state B. Now what is
the total number of configurations WB of the system?
( N + M )!
WB =
n!m!( N + M n m)!
Now take N=M and n=m for the following two problems:
(c) Using Stirlings approximation on page 56, what is the ratio WA/WB?
2N
2 N 2n
N 2 N 2n
2

W A
N
!

(n!) 2 (2
N 2n)!
e

e
2
2 N 2 n

2
2 n
=

2N
2 N 2n
N
(2 N )!
WB
n!( N n)!
2
2N N n

e

e

(d) Which state, A or B has the greatest entropy? Calculate the entropy difference
given by: S = SA SB = kln(WA/WB).

W

S = S A S B = k log A k log 2 2 n 2nk log 2 < 0

WB
so SA < SB.
6.10 Maximum entropy for a three-sided die. You have a three-sided die, with
numbers 1,2 and 3 on the sides. (So the values we can observe are 1=1, 2=2, 3=3. For a
series of N dice rolls, you observe an average score <> per roll using the maximum
entropy principle.
(a) Write expressions that show how to compute the relative probabilities of
occurrence of the three sides, n1*/N, n2*/N and n3*/N, if is given.
Like the example in the book, we can use the Boltzmann probabilities to maximize the
e i
entropy. So the general form for each probability is: pi = 3
e i
i =1

To simplify, take x = e so p1 =

, p2 =

x
x3
,
.
p
=
3
x + x2 + x3
x + x2 + x3

x + x2 + x3
3
x + 2 x 2 + 3x 3
We know that: =

i p i so =
x + x2 + x3
i =1
We can solve this for x in terms of by using the quadratic equation:

x=

2 + (2 ) 2 4(3 )(1 ) 2 + 3 2 + 12 8
=
2(3 )
2(3 )
To get the probability of turning up a 1, 2, or 3, we would first calculate x and then
plug back in to the equations for p1, p2, and p3.
(b) Compute n1*/N, n2*/N and n3*/N if =2.
x=1 so p1=p2=p3 = 1/3.
(c) Compute n1*/N, n2*/N and n3*/N if =1.5.
13 1
x=
0.4343 so p1= 0.6162 p2= 0.2627 and p3 = 0.1162
2
(d) Compute n1*/N, n2*/N and n3*/N if =2.5.
13 1
x=
2.3028 so p1= 0.1162 p2= 0.2627 and p3 = 0.6162.
6

1 of 4 pages
NAME_____________________________________
BE.011/2.772
Exam 2
19 March 2004
Conversions & Constants
k (boltzman constant) = 1.38017 x 10-23 J/K

Gas Constant R = Nk = 8.3144J/K-mol

Avagadros Number = 6.022 x 1023

1 atm = 101,325 N/m2

1 N = 1 J/m

1 J = 0.23901 cal

1. Home soda seltzer makers were popular in the Czech Republic during the early-mid 20th
century, and can now be purchased from various high-end purveyors of kitchenware. To
make seltzer, a heavy-duty bottle (to withstand pressure) is filled with water, leaving a
little air space on top. A charger/dispensing apparatus is screwed on to make a tight seal.
A metal canister containing 10 cm3 of compressed CO2 is placed in a holder and screwed
onto the charger/dispensing apparatus, piercing a metal membrane on the CO2 canister
and releasing CO2 into the bottle, where approx 50cm3 air space is available.
a. If the process occurs adiabatically, as expansion of the gas into the new
volume, what is the temperature of the gas at the end of the process? You
may consider that the air initially present in the bottle does not contribute to the
final state; i.e., you may consider this adiabatic expansion of the CO2 into a new
volume of 60 cm3. The constant pressure heat capacity of CO2 at the starting
temperature, room temperature (25C) is Cp = 37.4 J/mol-K. You can assume this
is an ideal gas, for which the molar heat capacity Cv = Cp R.

2 of 4 pages
b. Frost forms on the metal canister. What does this indicate about the assumption
that the process is adiabatic? How much energy would be required to cool the
canister from room temperature (25C) to 10C and form ~1 gm (~0.05 mols) of
ice if the heat capacity of the canister (mass x Cp) is 4300J/k and the heat of
fusion of ice Hmelt ~ 6000 J/mol (you can neglect the heat associated with
cooling the water vapor to 10C). How does that compare to what the enthalpy
change would be for cooling CO2 from room temperature to the temperature you
calculated in part a, if the cooling were done at constant pressure and the total
number of moles of CO2 is 0.2?

2. For the time being, it is still legal to Supersize your meal at MacDonalds. You order,
and eat, the following SuperSize meal (data obtained from the MacDonalds web site)
Double Quarter Pounder with Cheese, 770 calories
Super Size French Fries, 610 calories
Chocolate Triple Thick Shake (32 fl oz cup), 1150 calories
Baked Apple Pie, 260 calories
total calories 2790
Through a freak accident, as soon as you finish the last morsel, you suddenly become an
adiabatic system. How much does your body temperature rise if all of the calories in the
meal are converted to heat? Note that what is reported as calories are actually kcal
(i.e., the total heat generated from the meal is 2790 kcal). For the calculation, estimate
your weight as 60 kg, and your average heat capacity Cp as 1.0 kcal/kg/K.

3 of 4 pages

3. Consider again the text example we discussed in class of collapse of a 4-mer polymer
chain in a poor solvent. In that example, a simple 2-D lattice model was used to build an
expression for the free energy of the collapsed and open state, where the adjacent
monomers interact with energy U= per monomer-monomer interaction. Free energy
arguments were then used to define the temperature T0 where half the chains are in the
collapsed state and half in the open state.
a. Does a 6-mer chain (see figure for possible open and collapsed configurations)
made of the same monomers have a higher or lower value of T0? Provide
convincing evidence of your answer, but you do not need to calculate the
precise value of T0. For these purpose, we can define a collapsed chain as
having one or more monomer-monomer interactions. We emphasize that you do
not need to provide the precise value, just convincing evidence for whether T0 is
higher or lower for the 6-mer chain compared to the 4-mer.
b. Can you generalize your answer for larger N?

4 of 4 pages

4. Chymotrypsinogen, secreted by the pancreas, is the precursor to chymotrypsin, a

digestive enzyme that acts in the small intestine (and is thus necessary for completion of
the activity described in problem 2). At 25C and pH 3, the enthalpy of denaturation of
chymotrypsinogen is H = 39 kcal/mol. At body temperature (37C), the denaturation
H = 8 kcal/mol (also at pH 3). What is the change in heat capacity of the protein upon
denaturation? You may presume that all experiments were conducted at 1 atm pressure.

5. Ice at 1 atm and 0C has a density of 0.917 gm/cm3. Water at the same T and P has a
density of 0.9998 gm/cm3. At 1 atm and 0C, Hmelt = 5.9176kJ/mol.
a. What is the entropy change on melting?
b. What is U during melting?

Page 1 of 6
Name:
BE.011/2772J
Spring 2004
QUIZ III
You have 1 hour for this exam.
CLOSED BOOK
3 pages notes allowed

1 (45 points)
2 (30 points)
3 (25 points)
total (100 points)
Some formulas, constants, and conversions you may need:
k = 1.38 10-23 J/K
NA = 6.02 1023
0C = 273K

Continued on next page....

Page 2 of 6
Name:
1.) DNA supercoiling
We saw that circular DNA plasmids can supercoil into right or left handed supercoils with
number of turns:

=-1

=0

=1

The energy in which it has turns is

= 2B
where B is a constant.
a) (10 pts) Draw the first few energy levels. Label the levels with their values of , and draw the
energy levels to scale as much as possible (or label with its value of ). Include degeneracies.

Page 3 of 6
Name:
b) (2 pts) What are the units of B?

c) (18 pts) If T = B/k, what is the population of each of the energy levels at equilibrium? Which
microstates are relevant (i.e., populated?). Levels which have less than 1% of the population can
be considered negligible. Make a plot of population vs. energy justifying this.

Continued on next page....

Page 4 of 6
Name:
d) (5 pts) Calculate average energy per plasmid, <> at T = B/k. Use the same cutoff as you did
in part c. Leave in terms of B.

e) (5 pts) AFM is a technique which images the supercoils of DNA on a surface. It can
distinguish the molecules from one another. Lets say we have 100 of these plasmids at a
temperature of T = B/k. Calculate the entropy per plasmid.

Page 5 of 6
Name:
f) (5 pts) Calculate the free energy F of the entire system. Use the truncated form of the partition
function you obtained in part c). Leave your answer in units of B.

2.) Protein folding

A 99 residue protein has only one folded -helical conformation.
a) (10 pts) For the reaction

helix

random coil

Calculate S assuming that each residue can have three different configurations. Is S >, <, or =
0? Does this make sense? Why?

Page 6 of 6
Name:
b) (10 pts) What is H required for melting point to be 50C? The melting point is defined as
when the equilibrium constant is equal to 1.

c) (10 pts) Suppose you do an experiment on a solution of this protein and find S<0. Does this
concur with what you found in part a)? Why do you think this is?

3.) (25 pts) Protein dimerization equilibrium

You are interested in the tendency of a certain protein to form dimers. You obtain data for the
equilibrium between monomer and dimer at 20C and 30C and then your equipment breaks.
Predict the equilibrium constant at 37C if you found that the values of K at 20C and 30 C were
K20 = 18,000M-1 and K30 = 20,000M-1.

BE.011 MATLAB Tutorial

March 17, 2003
Originally written by Nate Tedford
for BE.420

How to Start and Run MATLAB

On a Mac or PC, run as you would any
other program..just point and click
All 12 PCs in the building 26
computer lab have MATLAB 6.5 (and
Solver!!) now installed

Running MATLAB on MIT server.

At the MIT server prompt, type:
MITserver% add matlab;
MITserver% matlab

First time, make a subdirectory:

MITserver% mkdir ~/matlab

In MATLAB, you can access additional MIT

server options by typing:
help MIT server
On some workstations, you can access newer
version of MATLAB by typing:
matlab-desktop at the matlab prompt

MATLAB Helpdesk
At the MATLAB prompt, type:
helpdesk

This will gives you a searchable command

help index which is toolbox specific and
more similar to the help resources that you
will see in the PC version 6.xx

Functions:

MATLAB Basics

Matrix formulation: fast calculation

Strong in numerical, but weak in analytical
General purpose math solvers: nonlinear equations, ODEs,
PDEs, optimization

Basic mode of usage:

Interactive mode
Permanent MATLAB files (M-files)
M-script
Functions

M-script and Functions must be written in separate

files
Note: M-files are saved in Work folder in the MATLAB
program files subdirectory

Basic Syntax
Case sensitive variable name
Library of Reserved Words
These will appear in blue if you are writing your
code as an M-file

End statements with a ;

Vector: Vec(i)
Matrix: Mat(i,j,)
Element by element matrix operations:
.*, ./, .^2
General matrix operations:
Cross product (*)

Syntax for Variable Assignment

Simple Variable
Type A = 4;

Vector
Type A = [1 2 3 4];

Matrix
Type A = [1 2 3 4; 5 6 7 8];

Variable Assignment Continued

Assignment of one value in a matrix
Type b = A(2,1); (same as b=5 here)

Incremental Vectors
Typing:
Z = (1:5) gives increments of 1
i.e. Z = [1 2 3 4 5]

Z = (1:3:7) gives increments of 3 between 1 and 7

i.e. Z = [1 4 7]

Two Important Points

If you do not put a semi-colon at the end of
the line, the result of the operation for that
line will be displayed when your program is
run => BE CAREFUL!!
Assignment vs. Equals: Important in Loops!
Assignment: a = b
Equals: a == b

Looping in MATLAB
For Loop
for I = 1:N
for J = 1:N
A(I,J) = 1/(I+J-1)
end
end
All Boolean expressions work
Less than: <, Greater than: >, Equal to: ==, Not
equal to: ~=, Less than or equal to: <=, Greater than
or equal to: >=.

Looping Continued
If Statement
if I == J
A(I,J) = 2;
elseif abs(I-J) == 1
A(I,J) = -1;
else
A(I,J) = 0;
end
As in C++, While loops can also be executed in MATLAB

Basic MATLAB Commands

Matlab commands

Functions and descriptions

help functionname

Matlab on-line help for functions

lookfor searchphrase

To find matlab function with descriptions

containing the search phrase
To list all variables currently used

who
size(matrix)
ones(m,n)

To identify the dimensionality of the

matrix (use length(vector) for a vector)
To create a unit matrix of size m x n

print depsc filename.ps To print an active plot (later use lpr to

print in MIT server)

Comments
You can write comments between and after
lines of code by typing % in front of your
message
You shold write your name and assignment
info on top of each program
Lastly, use comments throughout the code
to show me that you know what youre
doing

Plotting Your Data

After you have called your function (and
assigned a variable name to the soln)
Type: figure (dont need a semicolon here)
Type: plot(t,X)
t is your time vector and X is the soln vector that you
named in your function call or part of your soln
matrix (i.e. X(:,1), first column of matrix)

Note: You can type semilogx(t,X) or

semilogy(t,X) to get a semilog plot of your
choosing

Plotting Your Data Continued..

Typing hold on after introducing a second
figure will allow you to plot multiple curves
on the same set of axes
Using the subplot(x,y,z), plot(t,X) sequence
will allow you to plot a matrix of graphs of
size (x,y) on the same page, with z being the
location of the graph in the matrix
Typing plot(t,X,letter) will allow you to
control the color of the line for that plot, type
help plot in prompt to see the color key

Labelling Axes,Making Legends

For the plot title, type:
title(title)

For the x and y axes, type:

xlabel(axis name)
ylabel(axis name)

To make a legend, type:

legend(name of curve 1,name of curve 2, etc.)

Type all of these commands after each figure

and plot command so that I know what you
are presenting in each graph!!

Saving Your Work

In the MATLAB prompt, type:
save filename

In Windows, just use the save icon or the save

option in the drop down menu under file
Make sure that your file is saved in the proper
directory so that it can run from the MATLAB
prompt
In MIT server this is the directory you named the first time
you ran MATLAB
In Windows, it is normally the Work folder

Running Your Saved Work

Type the name of the M-file in the Matlab
prompt and hit enter
Also make sure that any functions that your
program uses are in the same directory as this
main M-file

If there are any errors in your code, they

will show up as messages in red text in the
prompt window

Some advice on getting help

USE THE HELP SEARCH TOOL
In MATLAB 6.xx, type:
help functionname

In the MIT server clusters, version 5.xx, use the helpdesk

option

Debug carefully
Write your code a little at a time
Use flags to see where errors are

If debugging is going nowhere, ask a friend to

check things out

MIT Help
Go to:

http://web.mit.edu/answers/www/matlab/
The Copy Tech also has printouts of basic
MATLAB commands and operations, you can
pick up a copy for free there

If you have a Pentium 4

and you have MATLAB Version
6.0
Go to:

http://www.mathworks.com/
Search for Pentium 4, Matlab version 6.0,

and youll be directed to a link that gives

you instructions to fix everything.

Monte Carlo
Solving problems using random number generation.
Extremely useful in solving noisy or analytically intractable problems!

measuring the
volume of the
Nile

Useful information you can get after

doing many trials of your system:
Mean:
Standard Deviation:
Integral evaluation:

Other stuff

P(xi) estimated from your trials as:

#sightings of xi / total # trials
num trials
mean

Get information about biological

systems!
Molecular Monte Carlo lets you explore many
interesting properties
thermally-averaged structures
molecular charge distributions
reaction rate constants
free energies
dielectric constants
heat capacities
phase transition temperatures
just about anything.

Easier things to try right now

We can use Monte Carlo to estimate a whole
probability distribution function that might
otherwise be very difficult to compute
Or, we can take a predefined distribution
function and randomly draw numbers from it
to learn things about the distribution

Generating random numbers

from a specific distribution
Almost always start by drawing random numbers
from a uniform distribution where all values
between 0 and 1 are equally likely
Many programming languages have such
random number generators built-in

MATLAB random number

generators
MATLAB has a ton of built-in random
number generators so that you dont have to
worry about the algorithms
A great demo of these generators is
randtool.m
You can try it out by typing randtool at the
matlab prompt in MIT server

Polymerase Chain Reaction: A brief

introduction
Denaturation:
Heat sample to 94oC to
denature the DNA

Annealing
Add primers
Reduce temp. to 54oC

Extension
Increase temp. to 72oC
Add dNTPs and thermophilic
polymerase (Taq)

And repeat!

Pcrsteps.gif

How can we represent PCR

statistically?
PCR is an example of a
branching process. During
each cycle of the reaction, we
can generate one or two
branches for each existing
copy of DNA (i.e. each strand
either gets copied or it
doesnt)

Scary model for PCR

Probability that PCR tube contains n1 copies given that we
started with n0:
n0 n1 n0

(1 )2 n0 n1
P (n1 | n0 { }) =
n1 n0
Apply Bayesian sum and product rule:
P( A | C ) = P( A | B) P( B | C )
B

And obtain an expression for the probability that after k

cycles the tube contains nk copies:
P(nk | n0 { }) = L P (nk | nk 1{ }) P (nk 1 | nk 2 { })L P(n1 | n0 { })
nk 1 nk 2

n1

Brownian motion and diffusion

Brownian motion: the constant and seemingly
random motion of particles due to the kinetic
energy imparted on one another as they collide
Diffusion: the macroscopic phenomenon by
which we observe the spreading out of
particles in space due to accumulated
Brownian motion
http://www.phys.virginia.edu/classes/109N/more_stuff/Applets/brownian/brownian.html

Brownian Motion
Not really randomif knew the location
and velocity of every particle we could
predict trajectories at any later time
Realistically this is impossible for system
with large number of particles
Describe by the random walk