Skip to content

okezue/ImpAgingSim

Repository files navigation

IMP Heteropolymer Aging Simulation (Metropolis MC)

Complete pipeline for simulating aging dynamics in the IMP heteropolymer with tunable disorder correlations.

Pipeline

Script Purpose
imp_aging.py Metropolis MC simulation, temperature quench, two-time observables
analyze_results.py Disorder-averaged timescales, aging exponent mu, collapse scores
extended_analysis.py KWW fitting, MSD/alpha2 analysis, energy/Rg tracking
quantitative_analysis.py Power-law fits, ensemble separation metrics
robustness_analysis.py Bootstrap CIs, multi-threshold tau, smoothed t*
create_comparison_figures.py Overlay figures for iid vs correlated
test_imp_aging.py 47-test suite (energy, disorder, observables, KWW)

Quickstart

python3 imp_aging.py \
  --out output/results.csv \
  --ensemble iid correlated \
  --epsilon 0 3 6 \
  --n_disorder 20 --n_traj 8 \
  --seed 123

python3 analyze_results.py --csv output/results.csv --outdir analysis --mono
python3 extended_analysis.py --csv output/results.csv --outdir analysis/extended
python3 -m pytest test_imp_aging.py -v

Observables

Observable Description
Q(t_w,t) Contact overlap (fraction of non-bonded pairs in contact at both t_w and t_w+t)
chi4 N_pairs * (var Q across trajectories) - dynamical heterogeneity
D4 Mean squared change in pairwise distances
MSD Per-bead mean-square displacement
alpha2 Non-Gaussianity parameter (3/5)*<r^4>/<r^2>^2 - 1
Energy Total potential energy at each snapshot
Rg Radius of gyration
n_contacts Number of non-bonded pairs in contact

Disorder Ensembles

  • iid: eta_ij ~ N(0,1) independent
  • correlated: eta_ij = kappa * sigma_i * sigma_j + sqrt(1-kappa^2) * xi_ij, where sigma is a Markov chain on {+1,-1} with persistence pi

About

Monte Carlo simulation of impatient spin glass aging dynamics - comparing IID vs correlated disorder ensembles

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors