implemented log likelihood method for pairwise interactions

src/dialect/models/interaction.py

@@ -28,9 +28,79 @@ def __init__(self, gene_a, gene_b):
     #                        Likelihood & Metric Evaluation                        #
     # ---------------------------------------------------------------------------- #
 
-    # Placeholder for methods to compute interaction-specific metrics (e.g., likelihood ratio)
-    # def compute_log_likelihood(self):
-    #     pass
+    def compute_log_likelihood(self, tau):
+        """
+        Compute the complete data log-likelihood for the interaction given the parameters tau.
+
+        The log-likelihood function is given by:
+            \sum_{i=1}^{N} \log ()
+                \mathbb{P}(P_i = c_i)\mathbb{P}(P_i' = c_i') \tau_{00} +
+                \mathbb{P}(P_i = c_i)\mathbb{P}(P_i' = c_i' - 1) \tau_{01} +
+                \mathbb{P}(P_i = c_i - 1)\mathbb{P}(P_i' = c_i') \tau_{10} +
+                \mathbb{P}(P_i = c_i - 1)\mathbb{P}(P_i' = c_i' - 1) \tau_{11}
+            )
+
+        where:
+            - `N` is the number of samples.
+            - `P_i` and `P_i'` represent the RVs for passenger mutations for the two genes.
+            - `c_i` and `c_i'` are the observed counts of somatic mutations for the two genes.
+            - `tau = (tau_00, tau_01, tau_10, tau_11)` are the interaction parameters.
+
+        :param tau (tuple): A tuple (tau_00, tau_01, tau_10, tau_11) representing the interaction parameters.
+        :return (float): The log-likelihood value.
+        :raises ValueError: If `bmr_pmf` or `counts` are not defined for either gene, or if `tau` is invalid.
+        """
+        if not self.gene1.bmr_pmf or not self.gene2.bmr_pmf:
+            raise ValueError("BMR PMFs are not defined for one or both genes.")
+        if not self.gene1.counts or not self.gene2.counts:
+            raise ValueError("Counts are not defined for one or both genes.")
+        if not all(
+            0 <= t <= 1 for t in [self.tau_00, self.tau_01, self.tau_10, self.tau_11]
+        ) or not np.isclose(
+            sum([self.tau_00, self.tau_01, self.tau_10, self.tau_11]), 1
+        ):
+            logging.info(
+                f"Invalid tau parameters: tau_00={self.tau_00}, tau_01={self.tau_01}, tau_10={self.tau_10}, tau_11={self.tau_11}"
+            )
+            raise ValueError(
+                "Invalid tau parameters. Ensure 0 <= tau_i <= 1 and sum(tau) == 1."
+            )
+
+        gene_a_counts = self.gene_a.counts
+        gene_b_counts = self.gene_b.counts
+
+        logging.info(f"Computing log likelihood for interaction {self.name}.")
+        logging.info(
+            f"tau_00: {self.tau_00}, tau_01: {self.tau_01}, tau_10: {self.tau_10}, tau_11: {self.tau_11}"
+        )
+
+        log_likelihood = sum(
+            np.log(
+                (  # \mathbb{P}(P_i = c_i)\mathbb{P}(P_i' = c_i') \tau_{00}
+                    self.gene_a.bmr_pmf.get(c_a, 0)
+                    * self.gene_b.bmr_pmf.get(c_b, 0)
+                    * self.tau_00
+                )
+                + (  # \mathbb{P}(P_i = c_i)\mathbb{P}(P_i' = c_i' - 1) \tau_{01}
+                    self.gene_a.bmr_pmf.get(c_a, 0)
+                    * self.gene_b.bmr_pmf.get(c_b - 1, 0)
+                    * self.tau_01
+                )
+                + (  # \mathbb{P}(P_i = c_i - 1)\mathbb{P}(P_i' = c_i') \tau_{10}
+                    self.gene_a.bmr_pmf.get(c_a - 1, 0)
+                    * self.gene_b.bmr_pmf.get(c_b, 0)
+                    * self.tau_10
+                )
+                + (  # \mathbb{P}(P_i = c_i - 1)\mathbb{P}(P_i' = c_i' - 1) \tau_{11}
+                    self.gene_a.bmr_pmf.get(c_a - 1, 0)
+                    * self.gene_b.bmr_pmf.get(c_b - 1, 0)
+                    * self.tau_11
+                )
+            )
+            for c_a, c_b in zip(gene_a_counts, gene_b_counts)
+        )
+        return log_likelihood
+
     #
     # def compute_likelihood_ratio(self):
     #     pass