new models commited

inst/stan/dgu_paired.stan

@@ -0,0 +1,149 @@
+functions {
+  real zibb_lpmf(int y, int n, real mu, real phi, real kappa) {
+    if (y == 0) {
+      return log_sum_exp(bernoulli_lpmf(1 | kappa),
+      bernoulli_lpmf(0 | kappa) +
+      beta_binomial_lpmf(0 | n, mu * phi, (1 - mu) * phi));
+    } else {
+      return bernoulli_lpmf(0 | kappa) +
+      beta_binomial_lpmf(y | n, mu * phi, (1 - mu) * phi);
+    }
+  }
+
+  int zibb_rng(int y, int n, real mu, real phi, real kappa) {
+    if (bernoulli_rng(kappa) == 1) {
+      return (0);
+    } else {
+      return (beta_binomial_rng(n, mu * phi, (1 - mu) * phi));
+    }
+  }
+
+  real z_rng(real a, real b, real zi) {
+    if (bernoulli_rng(zi) == 1) {
+      return (0);
+    } else {
+      return(inv_logit(a+b));
+    }
+  }
+}
+
+data {
+  int<lower=0> N_sample;                   // number of repertoires 
+  int<lower=0> N_gene;                     // gene
+  int<lower=0> N_individual;               // number of individuals
+  int<lower=0> N_condition;                // number of conditions
+  array [N_individual] int N;              // number of tries
+  array [N_gene, N_individual] int Y;      // number of heads for each coin
+  array [N_individual] int condition_id;   // id of conditions
+  array [N_sample] int individual_id;      // id of replicate
+}
+
+transformed data {
+  // convert int to real N for easier division in generated quantities block
+  array [N_individual] real Nr;
+  Nr = N;
+}
+
+parameters {
+  real <lower=0> phi;
+  real <lower=0, upper=1> kappa;
+
+  vector [N_gene] alpha;
+
+  vector <lower=0> [N_condition] sigma_condition;
+  vector <lower=0> [N_condition] sigma_individual;
+  real <lower=0> sigma_alpha;
+
+  array [N_individual] vector [N_gene] z_alpha_individual;
+  array [N_individual] vector [N_gene] z_beta_individual;
+  array [N_condition] vector [N_gene] z_beta_condition;
+}
+
+transformed parameters {
+  array [N_individual] vector <lower=0, upper=1> [N_gene] theta;
+  array [N_individual] vector [N_gene] alpha_individual;
+  array [N_individual] vector [N_gene] beta_individual;
+  array [N_condition] vector [N_gene] beta_condition;
+
+  for(i in 1:N_condition) {
+    beta_condition[i] = 0 + sigma_condition[i] * z_beta_condition[i];
+  }
+
+  for(i in 1:N_individual) {
+    alpha_individual[i] = alpha + sigma_alpha * z_alpha_individual[i];
+    beta_individual[i]  = beta_condition[condition_id[i]] + sigma_individual[condition_id[i]] * z_beta_individual[i];
+  }
+
+  for(i in 1:N_sample) {
+    theta[i] = inv_logit(alpha_individual[individual_id[i]] + beta_individual[individual_id[i]]);
+  }
+}
+
+model {
+  target += beta_lpdf(kappa | 1.0, 5.0);
+  target += exponential_lpdf(phi | 0.01);
+  target += normal_lpdf(alpha | -3.0, 3.0);
+
+  for(i in 1:N_condition) {
+    target += std_normal_lpdf(z_beta_condition[i]);
+  }
+  for(i in 1:N_individual) {
+    target += std_normal_lpdf(z_beta_individual[i]);
+  }
+
+  target += cauchy_lpdf(sigma_individual | 0.0, 1.0);
+  target += cauchy_lpdf(sigma_condition | 0.0, 1.0);
+  target += cauchy_lpdf(sigma_alpha | 0.0, 1.0);
+
+  for(i in 1:N_individual) {
+    for(j in 1:N_gene) {
+      target += zibb_lpmf(Y[j,i] | N[i], theta[i][j], phi, kappa);
+    }
+  }
+}
+
+generated quantities {
+  // PPC: count usage (repertoire-level)
+  array [N_gene, N_individual] int Yhat_rep;
+
+  // PPC: proportion usage (repertoire-level)
+  array [N_gene, N_individual] real Yhat_rep_prop;
+
+  // PPC: proportion usage at a gene level in condition
+  array [N_condition] vector [N_gene] Yhat_condition_prop;
+
+  // LOG-LIK
+  array [N_individual] vector [N_gene] log_lik;
+
+  // DGU matrix
+  matrix [N_gene, N_condition*(N_condition-1)/2] dgu;
+  matrix [N_gene, N_condition*(N_condition-1)/2] dgu_prob;
+  int c = 1;
+
+  //TODO: speedup, run in C++ not big factor on performance
+  for(j in 1:N_gene) {
+    for(i in 1:N_individual) {
+      Yhat_rep[j, i] = zibb_rng(Y[j, i], N[i], theta[i][j], phi, kappa);
+      log_lik[i][j] = zibb_lpmf(Y[j, i] | N[i], theta[i][j], phi, kappa);
+
+      if(Nr[i] == 0.0) {
+        Yhat_rep_prop[j, i] = 0;
+      }
+      else {
+        Yhat_rep_prop[j, i] = Yhat_rep[j,i]/Nr[i];
+      }
+    }
+    for(g in 1:N_condition) {
+      Yhat_condition_prop[g][j] = z_rng(alpha[j], beta_condition[g][j], 0);
+    }
+  }
+
+  // DGU analysis
+  for(i in 1:(N_condition-1)) {
+    for(j in (i+1):N_condition) {
+      dgu[,c] = beta_condition[i]-beta_condition[j];
+      dgu_prob[,c]=to_vector(Yhat_condition_prop[i])-to_vector(Yhat_condition_prop[j]);
+      c = c + 1;
+    }
+  }
+}

inst/stan/dgu_paired_rep.stan

@@ -0,0 +1,165 @@
+functions {
+  real zibb_lpmf(int y, int n, real mu, real phi, real kappa) {
+    if (y == 0) {
+      return log_sum_exp(bernoulli_lpmf(1 | kappa),
+      bernoulli_lpmf(0 | kappa) +
+      beta_binomial_lpmf(0 | n, mu * phi, (1 - mu) * phi));
+    } else {
+      return bernoulli_lpmf(0 | kappa) +
+      beta_binomial_lpmf(y | n, mu * phi, (1 - mu) * phi);
+    }
+  }
+
+  int zibb_rng(int y, int n, real mu, real phi, real kappa) {
+    if (bernoulli_rng(kappa) == 1) {
+      return (0);
+    } else {
+      return (beta_binomial_rng(n, mu * phi, (1 - mu) * phi));
+    }
+  }
+
+  real z_rng(real a, real b, real zi) {
+    if (bernoulli_rng(zi) == 1) {
+      return (0);
+    } else {
+      return(inv_logit(a+b));
+    }
+  }
+}
+
+data {
+  int<lower=0> N_sample;                   // number of repertoires 
+  int<lower=0> N_gene;                     // gene
+  int<lower=0> N_individual;               // number of individuals
+  int<lower=0> N_condition;                // number of conditions
+  int<lower=0> N_replicate;                // number of replicates
+  array [N_individual] int N;              // number of tries
+  array [N_gene, N_individual] int Y;      // number of heads for each coin
+  array [N_individual] int condition_id;   // id of conditions
+  array [N_sample] int individual_id;      // id of individual
+  array [N_sample] int replicate_id;       // id of replicate
+}
+
+transformed data {
+  // convert int to real N for easier division in generated quantities block
+  array [N_individual] real Nr;
+  Nr = N;
+}
+
+parameters {
+  real <lower=0> phi;
+  real <lower=0, upper=1> kappa;
+
+  vector [N_gene] alpha;
+
+  vector <lower=0> [N_condition] sigma_condition;
+  vector <lower=0> [N_condition] sigma_individual;
+  real <lower=0> sigma_alpha;
+  real <lower=0> sigma_alpha_rep;
+  real <lower=0> sigma_beta_rep;
+
+  array [N_individual] vector [N_gene] z_alpha_individual;
+  array [N_individual] vector [N_gene] z_beta_individual;
+  array [N_condition] vector [N_gene] z_beta_condition;
+  array [N_individual, N_replicate] vector [N_gene] z_alpha_sample;
+  array [N_individual, N_replicate] vector [N_gene] z_beta_sample;
+}
+
+transformed parameters {
+  array [N_condition] vector [N_gene] beta_condition;
+  array [N_individual] vector [N_gene] alpha_individual;
+  array [N_individual] vector [N_gene] beta_individual;
+  array [N_individual, N_replicate] vector [N_gene] alpha_sample;
+  array [N_individual, N_replicate] vector [N_gene] beta_sample;
+  array [N_individual] vector <lower=0, upper=1> [N_gene] theta;
+
+  for(i in 1:N_condition) {
+    beta_condition[i] = 0 + sigma_condition[i] * z_beta_condition[i];
+  }
+
+  for(i in 1:N_individual) {
+    alpha_individual[i] = alpha + sigma_alpha * z_alpha_individual[i];
+    beta_individual[i]  = beta_condition[condition_id[i]] + sigma_individual[condition_id[i]] * z_beta_individual[i];
+  }
+
+  for(i in 1:N_sample) {
+    alpha_sample[individual_id[i], replicate_id[i]] = alpha_individual[individual_id[i]] + sigma_alpha_rep * z_alpha_sample[individual_id[i], replicate_id[i]];
+    beta_sample[individual_id[i], replicate_id[i]] = beta_individual[individual_id[i]] + sigma_beta_rep * z_beta_sample[individual_id[i], replicate_id[i]];
+    theta[i] = inv_logit(alpha_sample[individual_id[i], replicate_id[i]] + beta_sample[individual_id[i], replicate_id[i]]);
+  }
+}
+
+model {
+  target += beta_lpdf(kappa | 1.0, 5.0);
+  target += exponential_lpdf(phi | 0.01);
+  target += normal_lpdf(alpha | -3.0, 3.0);
+
+  for(i in 1:N_condition) {
+    target += std_normal_lpdf(z_beta_condition[i]);
+  }
+  for(i in 1:N_individual) {
+    target += std_normal_lpdf(z_beta_individual[i]);
+  }
+  for(i in 1:N_sample) {
+    target += std_normal_lpdf(z_alpha_sample[individual_id[i], replicate_id[i]]);
+    target += std_normal_lpdf(z_beta_sample[individual_id[i], replicate_id[i]]);
+  }
+
+  target += cauchy_lpdf(sigma_individual | 0.0, 1.0);
+  target += cauchy_lpdf(sigma_condition | 0.0, 1.0);
+  target += cauchy_lpdf(sigma_alpha | 0.0, 1.0);
+  target += cauchy_lpdf(sigma_alpha_rep | 0.0, 1.0);
+  target += cauchy_lpdf(sigma_beta_rep | 0.0, 1.0);
+
+  for(i in 1:N_individual) {
+    for(j in 1:N_gene) {
+      target += zibb_lpmf(Y[j,i] | N[i], theta[i][j], phi, kappa);
+    }
+  }
+}
+
+generated quantities {
+  // PPC: count usage (repertoire-level)
+  array [N_gene, N_individual] int Yhat_rep;
+
+  // PPC: proportion usage (repertoire-level)
+  array [N_gene, N_individual] real Yhat_rep_prop;
+
+  // PPC: proportion usage at a gene level in condition
+  array [N_condition] vector [N_gene] Yhat_condition_prop;
+
+  // LOG-LIK
+  array [N_individual] vector [N_gene] log_lik;
+
+  // DGU matrix
+  matrix [N_gene, N_condition*(N_condition-1)/2] dgu;
+  matrix [N_gene, N_condition*(N_condition-1)/2] dgu_prob;
+  int c = 1;
+
+  //TODO: speedup, run in C++ not big factor on performance
+  for(j in 1:N_gene) {
+    for(i in 1:N_individual) {
+      Yhat_rep[j, i] = zibb_rng(Y[j, i], N[i], theta[i][j], phi, kappa);
+      log_lik[i][j] = zibb_lpmf(Y[j, i] | N[i], theta[i][j], phi, kappa);
+
+      if(Nr[i] == 0.0) {
+        Yhat_rep_prop[j, i] = 0;
+      }
+      else {
+        Yhat_rep_prop[j, i] = Yhat_rep[j,i]/Nr[i];
+      }
+    }
+    for(g in 1:N_condition) {
+      Yhat_condition_prop[g][j] = z_rng(alpha[j], beta_condition[g][j], 0);
+    }
+  }
+
+  // DGU analysis
+  for(i in 1:(N_condition-1)) {
+    for(j in (i+1):N_condition) {
+      dgu[,c] = beta_condition[i]-beta_condition[j];
+      dgu_prob[,c]=to_vector(Yhat_condition_prop[i])-to_vector(Yhat_condition_prop[j]);
+      c = c + 1;
+    }
+  }
+}

man/d_zibb_4.Rd

@@ -0,0 +1,40 @@
+\name{d_zibb_4}
+\alias{d_zibb_4}
+\docType{data}
+\title{Simulated Ig gene usage data}
+
+\description{
+A small example dataset that has the following features:
+
+  \itemize{
+    \item 3 conditions
+    \item 3 samples per condition
+    \item 3 replicates per sample
+    \item 15 Ig genes
+  }
+This dataset was simulated from zero-inflated beta-binomial (ZIBB) 
+distribution. Simulation code is available in inst/scripts/d_zibb_4.R
+}
+
+\usage{
+data("d_zibb_4", package = "IgGeneUsage")
+}
+
+\format{
+A data frame with 4 columns:  
+\itemize{
+  \item "individual_id"
+  \item "condition"
+  \item "gene_name"
+  \item "gene_name_count"
+}
+This format is accepted by IgGeneUsage.
+}
+\source{
+Simulation code is provided in inst/scripts/d_zibb_4.R
+}
+\examples{
+data("d_zibb_4", package = "IgGeneUsage")
+head(d_zibb_4)
+}
+\keyword{d_zibb_4}