Simulation_dataset.benchmark_stats: fixed bug

app/lmm_benchmark.ml

@@ -15,7 +15,7 @@ let besnard2009 = {
   label = "besnard2009" ;
   tree = Bistro.Workflow.input "data/besnard2009/besnard2009.nhx" ;
   rooted = true ;
-  branch_scale = 10. ;
+  branch_scale = 1. ;
   ne_s = 4., 4. ;
 }
 
@@ -23,7 +23,7 @@ let oneline_rodent = {
   label = "online_rodent" ;
   tree = Bistro.Workflow.input "data/online_rodent/online_rodent.nhx"  ;
   rooted = true ;
-  branch_scale = 10. ;
+  branch_scale = 1. ;
   ne_s = 4., 4. ;
 }
 
@@ -31,7 +31,7 @@ let rubisco = {
   label = "rubisco" ;
   tree = Rubisco_dataset.(Path "data/rubisco" |> Query.tree ~branch_length_unit:`Amino_acid) ;
   rooted = false ;
-  branch_scale = 10. ;
+  branch_scale = 1. ;
   ne_s = 4., 4. ;
 }
 
@@ -44,7 +44,7 @@ let orthomam_echolocation = {
       (Codepitk.Orthomam_db.make "omm")
   ) ;
   rooted = false ;
-  branch_scale = 10. ;
+  branch_scale = 1. ;
   ne_s = 4., 4. ;
 }
 

lib/simulation_dataset.ml

@@ -212,6 +212,7 @@ let%pworkflow benchmark_statistics simulation ~labels ~results =
   let open Codepitk in
   let open OCamlR_base in
   let open Codepitk.Simulator.Site_independent_mutsel in
+  let module Codon = Codon.Universal_genetic_code.NS in
   let load_results fn col =
     let open Codepitk in
     let df = Dataframe.from_file fn  |> Rresult.R.failwith_error_msg in
@@ -233,7 +234,6 @@ let%pworkflow benchmark_statistics simulation ~labels ~results =
       l, `Numeric (Numeric.of_array_opt r)
     ) in
   let amino_acid_vector_of_codon_vector xs =
-    let module Codon = Codon.Universal_genetic_code.NS in
     Amino_acid.Vector.init (fun aa ->
         List.fold Codon.all ~init:0. ~f:(fun acc c ->
             if Amino_acid.equal aa (Codon.aa_of_codon c) then
@@ -258,7 +258,13 @@ let%pworkflow benchmark_statistics simulation ~labels ~results =
     let t =
       Amino_acid.Table.init (fun aa ->
           let aa = Amino_acid.to_char aa in
-          List.count seqs ~f:(fun s -> Char.equal (s : Dna.t :> string).[i] aa)
+          List.count seqs ~f:(fun s ->
+              let codon_str = String.sub (s : Dna.t :> string) ~pos:(i * 3) ~len:3 in
+              let codon = match Codon.of_string codon_str with
+                | Some c -> c
+                | None -> assert false
+              in
+              Char.equal (Amino_acid.to_char (Codon.aa_of_codon codon)) aa)
         )
     in
     (t :> int array)