% Test the Bayesian and asymptotic error estimates for BAR by performing many replications of an experiment and determining fraction of time the true error is within the anticipated error. clear; % PARAMETERS Nmax = 10; % maximum number of samples N_f + N_r nreplicates = 1000; % number of replications of the experiment to perform for each combination of (N_f,N_r) cis = [0.68, 0.95]; % confidence intervals at which to evaluate error % Harmonic oscillator is given by potential % U(x) = (K / 2) (x - x_0)^2 % probability density function % p(x) = (1/Z) exp[-\beta U(x)] % % Z = \int dx \exp[-\beta U(x)] % = \int dx \exp[-(x - x_0)^2 / (2 sigma^2)] % = sqrt(2 pi) sigma % F = - ln Z = - (1/2) ln (2 pi) - ln sigma % sigma^2 = (\beta K)^{-1} x_0 = 0.0; % equilibrium spring position for initial state K_0 = 1.0; % spring constant of initial state x_1 = 1.5; % equilibrium spring position for final state K_1 = 4.0; % spring constant of final state beta = 1.0; % inverse temperature % Determine number of confidence intervals to evaluate ncis = length(cis); % Compute Gaussian widths for harmonic oscillators sigma_0 = 1 / sqrt(beta * K_0); sigma_1 = 1 / sqrt(beta * K_1); % Define instantaneous work functions WF = @(x) (K_1/2)*(x-x_1).^2 - (K_0/2)*(x-x_0).^2; % work from state 0 -> 1 WR = @(x) (K_0/2)*(x-x_0).^2 - (K_1/2)*(x-x_1).^2; % work from state 1 -> 0 % Compute true free energy difference. true_df = - log(sigma_1) + log(sigma_0); % Plot system. plot_system; return % Estimate error distributions for each combination of (N_f,N_r) Pfrs_bayesian = zeros([Nmax+1, Nmax+1, ncis]); % Pfrs_bayesian(N_f,N_r,c) is the probability that the true error is within Bayesian confidence interval of cis(c) for N_f forward and N_r reverse samples Pfrs_asymptotic = zeros([Nmax+1, Nmax+1, ncis]); mask = zeros([Nmax+1, Nmax+1]); BAR_ML_bias_fr = zeros([Nmax+1, Nmax+1]); BAR_mean_bias_fr = zeros([Nmax+1, Nmax+1]); df_ML = zeros([nreplicates,1]); df_mean = zeros([nreplicates,1]); for N_f = 0:Nmax for N_r = 0:Nmax % if (N_f > 0) && (N_r > 0) && (N_f + N_r <= Nmax) % only evaluate those cases where there are some, but not too many, samples % if (N_f > 0) && (N_r > 0) && (N_f + N_r <= Nmax) % only evaluate those cases where there are some, but not too many, samples if (N_f + N_r > 1) mask(N_f+1, N_r+1) = 1; disp(sprintf('N_f = %d, N_r = %d', N_f, N_r)); % perform a number of replicates Nasymptotic = zeros([nreplicates, ncis]); Nbayesian = zeros([nreplicates, ncis]); for replicate = 1:nreplicates % perform a replicate of the experiment with fixed number of forward and reverse trajectories % Draw samples from stationary distributions at inverse temperature beta. x_f = sigma_0 * randn([N_f, 1]) + x_0; % samples from state 0 x_r = sigma_1 * randn([N_r, 1]) + x_1; % samples from state 1 % Compute forward and reverse work values. w_f = WF(x_f); w_r = WR(x_r); % compute confidence interval for asymptotic covariance estimate [df, ddf] = MBAR(w_f, w_r); % store statistics df_ML(replicate) = df; ddf_ML(replicate) = ddf; % Compute mean by Bayesian BAR. [df_low, df_this_mean, df_high] = BBAR(w_f, w_r, cis(1), df, ddf); df_mean(replicate) = df_this_mean; % compute free energy confidence intervals for c = 1:ncis % get confidence interval cutoff ci = cis(c); % compute confidence interval for Bayesian method [df_low, df_this_mean, df_high] = BBAR(w_f, w_r, ci, df, ddf); % record whether this was in the allowed region if (df_low <= true_df) && (true_df <= df_high) Nbayesian(replicate, c) = Nbayesian(replicate, c) + 1; end % compute normal confidence interval df_ci_lower = df - sqrt(2)*ddf*erfinv(ci); df_ci_upper = df + sqrt(2)*ddf*erfinv(ci); % DEBUG %disp(sprintf('MBAR: df = %8.3f +- %8.3f : [ %8.3f (%8.3f) %8.3f ]', df, ddf, df_ci_lower, true_df, df_ci_upper)); % record whether this was in the allowed region if (df_ci_lower <= true_df) && (true_df <= df_ci_upper) Nasymptotic(replicate, c) = Nasymptotic(replicate, c) + 1; end end % Compute fraction of time true free energy was within various confidence intervals. for c = 1:ncis P = sum(Nasymptotic(:,c)) / nreplicates; Pfrs_asymptotic(N_f+1, N_r+1, c) = P; dPfrs_asymptotic(N_f+1, N_r+1, c) = P * (1-P) / nreplicates; P = sum(Nbayesian(:,c)) / nreplicates; Pfrs_bayesian(N_f+1, N_r+1, c) = P; dPfrs_bayesian(N_f+1, N_r+1, c) = P * (1-P) / nreplicates; end % Estimate bias. BAR_ML_bias_fr(N_f+1, N_r+1) = mean(df_ML) - true_df; BAR_mean_bias_fr(N_f+1, N_r+1) = mean(df_mean) - true_df; end end end end % save all data save bar-test.mat % plot data generate_plot