An Elementary Mode Coupling Theory of Random Heteropolymer Dynamics
Extracted source text ยท arXiv cond-mat/9701099
\documentstyle[prl,aps,preprint]{revtex}
%\documentstyle[pnas]{article}
%\documentstyle[titlepage,12pt]{article}
%\setlength{\topskip}{0.0cm}
%\setlength{\textwidth}{17.0cm}
%\setlength{\textheight}{21.0cm}
%\setlength{\oddsidemargin}{0.cm}
%\setlength{\evensidemargin}{0.cm}
%\setlength{\topmargin}{-1.0cm}
%************************************************************************
\begin {document}
%\setlength{\baselineskip}{0.75cm}
%\twocolumn[
\draft
\title{An elementary mode coupling theory of random heteropolymer dynamics\\
\small{(glass transition / protein folding / Langevin equation )}\\
\small{Classification: Physics, Biophysics}
}
\author{Shoji Takada, John J. Portman, and Peter G. Wolynes}
\address{School of Chemical Sciences and Department of Physics, University of Illinois, Urbana, IL, 61801\\
{\it Contributed by Peter G. Wolynes} \\
Abbreviation: MCT, mode coupling theory ; RHP, random heteropolymer ; \\
DFI, dynamic functional integral.
}
\date{\today}
\maketitle
%]
\begin{abstract}
{The Langevin dynamics of a random heteropolymer and its dynamic glass transition
are studied using elementary mode coupling theory.
Contrary to recent reports using a similar framework,
a discontinuous ergodic-nonergodic phase
transition is predicted for all Rouse modes at a finite temperature $T_{\rm A}$.
For sufficiently long chains, $T_{\rm A}$ is almost independent of chain
length and is in good agreement with the value previously estimated by
a static replica theory.}
\end{abstract}
\pacs{}
%\vspace{0.5cm}
%\hrule
%\vspace{0.5cm}
\noindent
The self-organization of evolved biopolymers such as foldable proteins
ultimately depends upon the chain dynamics of heteropolymers.
The modern statistical mechanical view using energy landscape ideas
focuses on the analogy of folding with phase transitions in finite
size systems and exploits to a large extent our understanding of the
quasi-thermodynamic and static features of both regular systems
and frustrated disordered systems such as spin glasses\cite{Bryngelson95}.
The connection with detailed analytical theories of the time dependence
of fluctuational motions of a randomly interacting chain has only recently
received attention\cite{Kinzelbach,Timoshenko,Roan,Thirumalai}.
For completely random heteropolymers (RHP) the quasi-thermodynamic analyses
\cite{Takada96,Takada97}
using replica techniques give results parallel to those for other systems
with random first order phase transitions such as the mean field Potts
spin glasses\cite{Kirkpatrick87}.
An important feature of theories of random first order transitions is
the presence of two transitions; one is static, while the other is dynamic
and generally occurs at a higher temperature. The dynamical
transition signals a crossover to a motional mechanism
involving {\em a}ctivated motions\cite{Kirkpatrick87,Kirkpatrick}
(We denote this
dynamical transition temperature $T_{\rm A}$).
For RHP's two different kinds of approximate
quasi-static theories based on replicas\cite{Takada96,Takada97} and on the
generalized random energy model\cite{Plotkin} that takes into account
correlation in the energy landscape yield these two transitions.
Replica based techniques also yield estimates for the free energy
barrier heights of activated motions between the two
characteristic temperatures
\cite{Takada96,Takada97}.
Purely dynamical theories based on mode coupling ideas generally yield
a transition in harmony with the quasi-static analysis of
the dynamical transition which is in some sense a spinodal.
While mode coupling theory (MCT) is perturbatively well defined for
spin systems with long range random interactions\cite{Fischer,Kirkpatrick},
there are various versions
developed in the theory of fluids\cite{Kawasaki} and polymers\cite{Schweizer} that
are forced to make uncontrolled approximations.
On the basis of two such calculations, the validity of the emerging
picture of the dynamics and the energy landscape of RHP's
have been questioned\cite{Roan}.
Roan and Shakhnovich derive a mode coupling theory of the RHP
and explicitly solve it for a polymer in a good solvent\cite{Roan}.
It is no surprise that an uncollapsed chain has no dynamical transition,
but the authors further claim that this is actually a structural feature of the
RHP mode coupling equations independent of the state of collapse.
Thirumalai et al.\cite{Thirumalai} derive another
set of mode coupling equations
for a somewhat different model that exhibits no static replica symmetry breaking\cite{Garel}
and conclude there is a dynamical transition with a transition temperature that
depends on the length scale of the motional mode considered\cite{Thirumalai}.
A numerical treatment of self-consistent equations for heteropolymer collapse,
on the other hand, does show evidence for a dynamic freezing transition
\cite{Timoshenko}.
The inconsistency of these results with each other and
in the first two cases with replica calculations is disturbing.
The technical intricacy of these mode coupling calculations is a barrier to
understanding their inconsistencies. We have therefore derived mode
coupling equations using `elementary' methods like those used decades ago
in the theory of critical phenomena\cite{Kawasaki}. The resulting equations differ
in some respects from those of the earlier workers but clearly yield
transitions in harmony with the quasi-static results.
These equations also lead to an understanding of how the dynamical freezing
depends on chain length and state of collapse as well as
how the dynamics varies
for different modes of chain motion.
We consider a standard Hamiltonian for $N$ interacting beads
\begin{equation}
H=k_B T\sum_{i=1}^N ({\bf r}_{i+1}-{\bf r}_i)^2 +1/2\sum_{i\ne j} b_{ij}u(\Delta {\bf r}_{ij})
+V_{\rm ex}-\sum {\bf h}_i(t)\cdot{\bf r}_i(t),
\end{equation}
where ${\bf r}_i$ are the bead locations,
$b_{ij}$ is chosen as Gaussian random with mean
$b_0$ and variance $b^2$, $\Delta {\bf r}_{ij}=
{\bf r}_i-{\bf r}_j$, and $u({\bf r})$ is the two-body
interaction chosen as a Gaussian $\exp -{\bf r}^2/\sigma^2$.
$V_{\rm ex}$ is the excluded volume term which usually contains three body
terms, but it is enough, at the level of the present analysis for phantom chain,
to replace it with an effective Gaussian confinement
term $V_{\rm conf}=k_BTB\sum {\bf r}_i^2$, with the constant $B$ chosen so that the radius
of gyration $R_g$ becomes the physically required value determined by the packing fraction,
$\eta$\cite{onB}.
${\bf h}_i$ is the external force introduced for convenience in the
derivation of the response function.
For technical simplicity, we adopt a ring polymer model where ${\bf r}_{N+1}={\bf r}_1$.
The Langevin equation for the beads is
\begin{equation}\label{eq:LE}
\Gamma^{-1}\partial_t{\bf r}_i =-\partial \beta H /\partial{\bf r}_i
+{\bf \xi}_i
\end{equation}
Here
${\bf \xi}_i(t)$ is a Gaussian random force
for which the first two moments are given by
$\langle {\bf \xi}_i(t)\rangle =0$ and
$\langle {\bf \xi}_i(t) {\bf \xi}_j(t^\prime)\rangle=2\Gamma^{-1}
\delta_{ij}\delta(t-t^\prime) {\bf 1}$ ({\bf 1} is the 3 by 3 unit matrix),
$\Gamma$ determines the microscopic time scale
(it will be set to unity) and $\beta=1/k_B T$ as usual.
Since our main interest is in collapsed states where hydrodynamic effects are less
important, we ignore them for simplicity although their inclusion is not very hard.
A more essential simplification of
our treatment is the neglect of
entanglement effects. An uncrossable chain will possess more friction
than the phantom chain treated here\cite{Schweizer}, modifying the dynamical transition
temperature.
A standard tool for dealing with the average over the quenched randomness in
the Langevin dynamics has been the dynamic functional integral (DFI)
formalism\cite{Fischer,Kirkpatrick}. Although this approach is generally
accepted for the infinite range spin model, direct application of the
formalism to RHP models is questionable for two
reasons:
1)The heteropolymer interaction is short range so that the mean
field approximation is not exact; in such a situation, there exists some ambiguity
with the DFI formalism in its use of steepest descents.
2)In contrast to spin models, the Jacobian appearing in the DFI
formalism of the heteropolymer model depends on the randomness.
This point was neglected in previous treatments.
The usual formal identity for averaging over randomness
in the Hamiltonian for a spin-glass model is no longer precisely
true. Thus, we take an alternative and simpler route here; we use a
perturbation theory in the randomness $b_{ij}$ along
with a self-consistent prescription that corresponds
to a particular resummation of higher order terms.
In the case of the infinite range $p$-spin model
(including the Sherrington-Kirkpatrick model), this procedure
still leads to exactly the same results as the DFI formalism\cite{Kirkpatrick}.
We sketch the perturbation scheme briefly since details
will be published elsewhere\cite{Portman}.
The Langevin equation, Eq (\ref{eq:LE}), can be expressed as
\begin{equation}\label{eq:PLE}
{\bf r}_{l}(t) = \sum_i G_{0,li}*({\bf \xi}_i + {\bf h_i}
-\beta/2{\partial\over\partial {\bf r}_i}
\sum_{kj}b_{kj} u\left( \Delta {\bf r}_{kj}) \right),
\end{equation}
where $*$ represents the time
convolution and $G_{0,ij}(t-t^\prime)$ is the
$0$-th order response function $\partial \langle {\bf r}_i^{(0)}(t)\rangle/\partial {\bf h}_j(t^\prime)$.
Inserting the $0$-th order solution ${\bf r}_l^{(0)}=\sum G_{0,li}* {\bf \xi}_i$
into the right hand side of eq.(\ref{eq:PLE}) gives the first order solution
${\bf r}_l^{(1)}= {\bf r}_l^{(0)}-\beta/ 2 \sum_i G_{0,li}*{\partial /\partial{\bf r}^{(0)}_i}
\sum_{kj} b_{kj} u\left(\Delta{\bf r}^{(0)}_{kj}\right)$.
We insert this once again into eq.(\ref{eq:PLE}), yielding the second order solution.
After taking an average over $b_{ij}$, we multiply the resulting equation
by $G_0^{-1}$ which yields
\begin{equation}\label{eq:r2}
\Gamma^{-1}\partial_t{\bf r}^{(2)}_i=-\sum K_{ij}
{\bf r}^{(2)}_j(t)
+\mu \sum_j \int_{-\infty}^{t}{\rm d} t^\prime
\Delta G_{0,ij}(t-t^\prime)\nabla\nabla u\left(\Delta{\bf r}^{(0)}_{ij}(t)\right)\cdot
\nabla u\left(\Delta{\bf r}^{(0)}_{ij}(t^\prime)\right)
+{\bf \xi}_i(t),
\end{equation}
where $K_{ij}$ is the harmonic constant matrix including the confinement term,
\newline$K_{ij}{\bf 1}\equiv\partial^2/\partial {\bf r}_i\partial {\bf r}_j
k_B T\left[ \sum ({\bf r}_{i+1}-{\bf r}_i)^2 + B\sum {\bf r}_i^2\right]$,
$\mu=(\beta b)^2$, and $\Delta G_{0,ij}=G_{0,ii}-G_{0,ij}-G_{0,ji}+G_{0,jj}$.
The random noise correlation can also be calculated by perturbation theory.
With the use of the relation $C_{lm}=\sum_{ij}G_{li} * \langle {\bf \phi}_i{\bf \phi}_j \rangle * G_{mj}$,
we can derive the expression for the correlations up to
second order, $C^{(2)}$, which includes
$\langle {\bf \phi}_i(t){\bf \phi}_j (t^\prime)\rangle^{(2)}$.
A part of higher order terms can be taken into account by
employing a self-consistent prescription.
To this end, we first expand the memory kernel of eq.(\ref{eq:r2})
in $\Delta {\bf r}_{ij}(t^\prime)$. Then,
in the spirit of Kawasaki's derivation of MCT
of critical dynamics, we replace ${\bf r}_i^{(0)}$ in the
right hand side by ${\bf r}_i$ and $G_0$ by the
perturbed response function, $G$, giving a renormalized
Langevin equation,
\begin{equation}\label{eq:RGLE}
\Gamma^{-1}{\rm d}{\bf r}_i/{\rm d}t=-\sum K_{ij}
{\bf r}_j(t)
+\mu \sum_j \int_{-\infty}^{t}{\rm d} t^\prime
\Delta G_{ij}(t-t^\prime){\cal C}_{ij}( t-t^\prime)
\Delta {\bf r}_{ij}(t^\prime)+{\bf \phi}_i(t),
\end{equation}
where
${\cal C}_{ij}=\langle \partial \left[ \nabla\nabla u\left( \Delta{\bf r}_{ij}(t)\right)\cdot
\nabla u\left(\Delta{\bf r}_{ij}(t^\prime)\right)\right]/
\partial \Delta {\bf r}_{ij}(t^\prime)\rangle$.
In deriving this equation we used the isotropic symmetry of the model.
The random force ${\bf \phi}_i(t)$ is also renormalized in the same way.
In the expression for $\langle {\bf \phi}_i(t){\bf \phi}_j (t^\prime)\rangle^{(2)}$,
replacement of ${\bf r}_i^{(0)}$ by ${\bf r}_i$ leads to the
colored noise correlation function,
\begin{equation}\label{eq:RGnoise}
\langle {\bf \phi}_i(t) {\bf \phi}_j(t^\prime) \rangle
=2\Gamma^{-1}{\bf 1}\delta_{ij}\delta(t-t^\prime)
+\mu\left[ \delta_{ij}\sum_k M_{ik}(t-t^\prime)-
M_{ij}(t-t^\prime)\right],
\end{equation}
where
$M_{ij}=\langle \nabla u\left( \Delta{\bf r}_{ij}(t)\right)
\nabla u\left(\Delta{\bf r}_{ij}(t^\prime)\right) \rangle$ is the force-force correlation
on different beads.
In calculating ${\cal C}$ and $M$ we approximate
the stochastic process as Gaussian. The explicit expression for $M$ becomes
$M_{ij}(t)=\left[(1+s\Delta C_{ij}(0))^2-(s \Delta C_{ij}
(t))^2\right]^{-5/2}\Delta C_{ij}(t)$, where
$\Delta C_{ij}=C_{ii}-C_{ij}-C_{ji}+C_{jj}$ and $s=2/\sigma^2$.
For the ergodic phase, with the time-translational symmetry and
the relation
$\left[\partial_t \Delta C_{ij}(t-t^\prime)\right]{\cal C}_{ij}(t-t^\prime)=\partial_tM_{ij}(t-t^\prime)$,
we can verify that the fluctuation-dissipation theorem indeed holds.
From Eqs (\ref{eq:RGLE}) and (\ref{eq:RGnoise}),
a straightforward generalization of
ref.\cite{Kirkpatrick}'s scheme for the $p$-spin model leads to
a closed set of integro-differential equations for the correlation
functions. In the present case,
due to the sequence-translational symmetry of ring polymers, it is
more transparent to write the equation in the Rouse mode representation.
Namely, the Fourier transform of the correlation function with respect to the
sequence satisfies the equation
\begin{equation}\label{eq:tprop}
\Gamma^{-1}\partial_t c_p(t)+c_p(t)/c_p(0)+\mu \int_0^t dt^\prime \partial_t c_p(t-t^\prime)
\left[ m_0(t^\prime)-m_p(t^\prime) \right]=0,
\end{equation}
where $c_p(t)={\rm FT}_{(i-j)\rightarrow p}[C_{ij}(t)]$ is the Rouse mode
correlation function and
$m_p(t)= {\rm FT}_{(i-j)\rightarrow p}\left[M_{ij}(t)\right]$ is a
nonlinear function of all of the correlations $\{ c_{p^\prime}\}$.
The equal time correlation function $c_p(0)$
obeys
\begin{equation}\label{eq:scc0}
c_p(0)^{-1}={\rm FT}_{(i-j)\rightarrow p}[K_{ij}]-\mu[m_0(0)-m_p(0)]\equiv \kappa_p.
\end{equation}
If we introduce the normalized Edwards-Anderson (EA) Rouse-Zimm order parameters
$\tilde{q}_p=\lim_{t\rightarrow\infty}c_p(t)/c_p(0)$,
the self-consistent equation for $\tilde{q}_p$ becomes
\begin{equation}\label{eq:scea}
\tilde{q}_p / (1-\tilde{q}_p)=(\mu / \kappa_p) \left[m_0(\infty)-m_p(\infty) \right]\equiv F_p,
\end{equation}
which is isomorphous to the equations for the MCT of structural glasses\cite{Gotze}.
Notice that if $F_p\ne 0$ the RZ modes have a static offset indicating
the trapping in a local minimum
and that all $p$ modes are coupled in evaluating
$m_p(\infty)$. Thus, clearly, the dynamic glass transition,
if any, should occur
simultaneously for all modes in this analysis.
In general, the dynamic glass transition can be obtained by two
steps. We first solve eq.(\ref{eq:scc0}) for $c_p(0)$ and then
look for a non-trivial solution of eq.(\ref{eq:scea}).
Assuming $c_p(0)$, a purely static quantity, does not exhibit
any singular behavior\cite{Kirkpatrick,Gotze},
we use an unperturbed confined Rouse value
(i.e.\, ${\rm FT}_{(i-j)\rightarrow p}[K_{ij}]$) for $c_p(0)$ here
avoiding the first step.
Fig.1 shows EA parameters $\tilde{q}_p$ as a function of $\mu$ with
$p=1, 50, 200$ and $512$ for a $1024$-mer with parameters $\sigma=1$ and $\eta=0.8$.
The EA parameters indeed exhibit a discontinuous (first order) transition at
a critical value denoted by $\mu_{\rm A}$ for all $p$ modes.
Increasing the chain length up to 32768, we numerically show that
$T_{\rm A}$ defined by $\mu_{\rm A}=(b/k_BT_{\rm A})^2$ converges to a finite
value $T_{\rm A}=0.3b$(see Fig.2), which is in good agreement with the value $0.292b$
estimated by the static replica theory for essentially the same model\cite{Takada97}.
This value depends little on the choice of $\eta$ and thus the agreement
with the replica result is quite robust;
with $\eta=0.6$ and $\eta=1.0$ we found
$T_{\rm A}=0.28b$ and $T_{\rm A}=0.32b$, respectively.
Scanning a wide range of chain lengths,$N$, the self-consistent equations
yield two different transitions depending on chain length and state of collapase (Fig.2).
For collapsed phase (e.g., $\eta=0.8$), with short chain length,
the EA parameters first achieve a small
nonzero value (of order $10^{-2}$) giving a weakly nonergodic phase at a
parameter value $\mu_c$ (dashed curve in Fig.2).
A stronger nonergodic phase,
in which EA parameters become of order unity ($\mu_{\rm A}$),
appears for collapsed state
with chain length longer than $128$ (solid curve in Fig.2).
$\mu_c$ and $\mu_{\rm A}$ coalesce at $N\sim 512$ above which no
weakly nonergodic phase exists.
Under $\Theta$ solvent conditions (i.e., $\eta=0$),
on the other hand, we do not find any second transition and
$\mu_c$ continues to increase with $N$.
Since the statistical dynamical theory presented here is inherently appropriate for sufficiently large
systems, we interpret the transition at $\mu_{\rm A}$ to correspond
to the dynamical transition
found by the replica approach\cite{Takada96,Takada97}.
The weak transition ($\mu_c$) found here only for short chains is
fragile and could be an artifact of the model
and we postpone its detailed interpretation.
Fig.3 plots $q_p=\lim_{t\rightarrow\infty}c_p(t )$ as a function of $p$ as well as
its inverse Fourier transformed frozen static displacement $Q_{i-j}$.
Since the $p=0$ mode
is purely diffusive, we drop this component.
Roughly, the frozen fluctuations $q_p$ and $Q_{i-j}$'s vary
with $p$ and $i-j$ in a manner quite similar to the equal time fluctuations
$c_p(0)$ and $C_{i-j}$, respectively. Although the frozen modes are
quite localized in the bead
representation, the long wavelength fluctuations corresponding to small $p$
are considerable.
Which mode is dominantly responsible for the transition?
The reduction theorem of G\"{o}tze\cite{Gotze} suggests that
within MCT only one
particular mode causes the instability of the nonergodic glass phase and
the eigenvector of the stability matrix
$S_{pp^\prime}=\partial F_p/\partial \tilde{q}_{p^\prime}$
with the largest eigenvalue corresponds to the most {\em dangerous} mode.
Dashed curves in Fig.3 show the right-eigenvector of this dangerous mode in
the Rouse and bead representation, $\nu_p$ and $\nu_{i-j}$,
have similar behavior
to $c_p(0)$ and $C_{i-j}$, respectively.
Eq.(\ref{eq:tprop}) can directly be integrated to get the explicit time dependence of $c_p(t)$.
As in the MCT for structural glasses, critical slowing down is
expected to be found as $T_{\rm A}$ is approached.
Details will be given elsewhere\cite{Portman}.
As mentioned above, the analysis of Roan and Shakhnovich yielded
no singular behavior in the relaxation spectrum.
There are several points in their analysis which differ from the present one
but primarily their analysis is a lowest order perturbation with
respect to $b_{ij}$. However, it is clear from our argument that
the self-consistent nature of the MC calculation is essential for
the dynamic glass transition
and obviously this is included at a higher order in perturbation.
This is a principal reason for their discordant result.
Thirumalai et al.'s approach is quite similar in spirit to ours,
but they analyze an uncoupled mode equation analogous to eq.(\ref{eq:tprop})
which yields a scale-dependent transition temperature.
We believe that it is more accurate to include the coupling of
different Rouse modes, forcing them to freeze simulutaneously.
The consistency with the static replica calculation buttresses the belief.
We emphasize again that
the dynamic glass transition studied here by elementary MCT is, in reality,
a crossover being
smeared out due to entropic droplets as discussed in refs.\cite{Kirkpatrick87,Takada97}.
Thus, $T_{\rm A}$ should not be viewed as a crisp phase transition, but
a characteristic temperature at which the nature of chain dynamics changes
qualitatively from renormalized free chain dynamics to activated escape
from traps.
Above $T_{\rm A}$,
the so-called {\em cage effect} due to the
mode coupling addressed here renormalizes the Rouse relaxation through internal
friction yielding slow dynamics
analogous to the $\alpha$ relaxation of supercooled liquids,
while below $T_{\rm A}$(but still above the static glass transition temperature)
escape from localized free energy minima
by (local) thermal activations, i.e.\ entropic droplets, controls the dynamics.
In applying these results to protein folding we need to consider that proteins
are not entirely random but have evolved to satisfy the
{\em minimal frustration principle}\cite{Bryngelson95}
so as to avoid being trapped in nonnative local minima.
The sequence is designed so that energetically favorable but structurally
incorrect amino acid contacts are
minimized.
Thus, effects of the random heteropolymer's interaction considered here are
superimposed on the flow of configurations through a global {\em funnel}
leading to the native structure\cite{Bryngelson95}.
Qualitatively, the folding transition temperature $T_{\rm F}$ should be
larger than the static glass transition $T_{\rm K}$ for fast folding and
may perhaps be even larger than $T_{\rm A}$. If this is the case,
folding can be viewed as a diffusion process in an order parameter space,
where slow dynamics renormalized by mode coupling controls the configurational
diffusion rate.
We thank D.\ Thirumalai and E.\ Shakhnovich for making us aware of
their referenced works prior to publication.
We also thank K.\ S.\ Schweizer for stimulating discussions.
S.T.\ is a Postdoctoral Fellow for Research Abroad of the Japan Society for
the Promotion of Science. J.J.P.\ is supported by NSF grant CHE
95-30680.
P.G.W.\ is supported by NIH grant PHS 5 R01 GM44557 and was a
Scholar-in-Residence at the Forgaty International Center for
Advanced Study in the Health Sciences,
National Institues of Health, Bethesda, U.\ S.\ A.\ when this
work was completed.
%\setlength{\baselineskip}{0.65cm}
%************************************************************************
%***********************************************************************
%\setlength{\baselineskip}{13pt}
\begin{thebibliography}{99}
%\bibitem{}
%, . {\bf }, ().
\bibitem{Bryngelson95}
Bryngelson, J. D., Onuchic, J. N., Socci, N. D., and Wolynes, P. G., (1995)
{\it PROTEINS: Struct. Funct. Genetics.} {\bf 21}, 167-195.
\bibitem{Kinzelbach}
For random directed polymers see Kinzelbach, H. and Horner, H., (1993)
{\it J. Phys. I France}
{\bf 3}, 1329-1357.
\bibitem{Timoshenko}
Timoshenko, E. G., Kuznetsov, Yu. A., and Dawson, K. A., (1996)
{\it Phys. Rev. E.} {\bf 54}, 4071-4086.
\bibitem{Roan}
Roan, J. -R. and Shakhnovich, E. I., (1996) {\it Phys. Rev. E.}
{\bf 54}, 5340-5357.
\bibitem{Thirumalai}
Thirumalai, D., Ashwin, V., and Bhattacharjee, J. K., (1996)
to appear {\it Phys. Rev. Lett.}
\bibitem{Takada96}
Takada, S. and Wolynes, P. G., submitted.
\bibitem{Takada97}
Takada, S. and Wolynes, P. G., submitted.
\bibitem{Kirkpatrick87}
Kirkpatrick, T. R. and Wolynes, P. G., (1987) {\it Phys. Rev. B.}
{\bf 36}, 8552-8564.
\bibitem{Kirkpatrick}
a)Kirkpatrick, T. R. and Thirumalai, D., (1987) {\it Phys. Rev. B.}
{\bf 36}, 5388-5397 ;
b)Crisanti, A., Horner, H., and Sommers, H.-J., (1993)
{\it Z. Phys. B.} {\bf 92}, 257-271.
\bibitem{Plotkin}
Wang, J., Plotkin, S., and Wolynes, P. G., (1997) {\it J. Phys. I France}
to appear.
\bibitem{Fischer}
Fischer, K. H. and Hertz, J. A., (1991) {\sl Spin Glasses}, (Cambridge Univ. Press, UK).
\bibitem{Kawasaki}
Kawasaki, K., (1976) in {\it Phase Transitions and Critical Phenomena}
Vol. 5a, edited by Domb, C. and Green, M. E., (Academic Press, N.Y.),
pp. 165-403.
\bibitem{Schweizer}
Schweizer, K. S., (1989) {\it J. Chem. Phys.} {\bf 91}, 5802-5821 ;
{\sl ibid}, 5822-5839 ; (1993) {\it Physica Scripta} {\bf T49},
99-106.
\bibitem{Garel}
Garel, T., Leibler, L., and Orland, H., (1994) {\it J. Phys. II France}
{\bf 4}, 2139-2148.
\bibitem{onB}
In globule states, $R_g^2=3/4 \sqrt{2/B}$\cite{Sasai90}. $R_g$ is chosen
by the relation $\eta 4\pi R_g^3/3=N\sigma^3$, where $\eta$ is
the packing fraction.
The explicit consideration of three body terms will be presented elsewhere\cite{Portman}.
\bibitem{Sasai90}
Sasai, M. and Wolynes, P. G., (1990) {\it Phys. Rev. Lett.}
{\bf 65}, 2740-2743 ; (1992) {\it Phys. Rev. A.}
{\bf 46}, 7979-7997.
\bibitem{Portman}
Portman, J. J., Takada, S., and Wolynes, P. G., in preparation.
\bibitem{Gotze}
G\"{o}tze, W., (1991)
in {\sl Liquids, Freezing and the Glass transition}, edited by Levesque, D.,
Hansen, J. P., and Zinn-Justin, J., (Elsevier, New York), pp. 287-504.
%************************************************************************
\end{thebibliography}
\par\noindent
\vfill
%************************************************************************
%\begin{figure}
%\caption
\noindent
{Figure 1: Edwards-Anderson order parameters $q_p$ for $p = 1, 50, 200,$ and
$512$ as
a function of $\mu=(\beta b)^2$. Parameters used are $N=1024$,
$\Gamma =1$, $\sigma=1$ and $\eta=0.8$.}
%\label{fg:qwrtmu}
%\end{figure}
%\begin{figure}
%\caption
\noindent
{Figure 2: $\mu_{\rm A}=(b/k_BT_{\rm A})^2$ and $\mu_c$ as a function of chain length $N$.
$\mu_{\rm A}$ (solid curve) and $\mu_C$ (dashed curve) are actually
computed at chain length $2^p$ with $p=5,\cdots,15$.
$\eta=0$ for $\Theta$ solvent and $\eta =0.6, 0.8,$ and $1.0$ for collapsed state.
Parameters used are $\Gamma=\sigma=1$.}
%\label{fg:muwrtn}
%\end{figure}
%\begin{figure}
%\caption
\noindent
{Figure 3: Edwards-Anderson order parameters $q_p$ and $Q_{i-j}$
as a function of $p$ and $i-j$ (solid curves) and
vector components $\nu_p$ and $\nu_{i-j}$ of the
dangereous mode in $p$ and $i-j$ representations (dashed curves).
Parameters used are
the same as Fig.1 and $\mu$ is slightly above $\mu_{\rm A}$.}
%\label{fg:qwrtp}
%\end{figure}
%************************************************************************
\parskip 20pt \par\noindent
\end{document}