notes: high conf prots: more stats

notes/gdb_mmseqs_highconf.md

@@ -13,19 +13,36 @@ df = pandas.read_csv("/scratch/users/vgalata/gdb/results/report/mmseqs2_highconf
 
 # proteins/protein clusters covering all assemblies
 df_all = df.loc[df["mmseqs2_all"],["tool_prot_id", "mmseqs2_cluster"]]
-# proteins/protein clusters NOT covering all assemblies but with ave. metaT cov. >= 10
+# proteins/protein clusters NOT covering all assemblies BUT with ave. metaT cov. >= 10
 df_cov = df.loc[~(df["mmseqs2_all"]) & (df["ave_cov"] >= 10),["tool_prot_id", "mmseqs2_cluster"]]
+# proteins/protein clusters covering all assemblies AND with ave. metaT cov. >= 10
+df_allcov = df.loc[(df["mmseqs2_all"]) & (df["ave_cov"] >= 10),["tool_prot_id", "mmseqs2_cluster"]]
 
 # print stats
+total_prots     = df.shape[0]
+total_clusters  = len(set(df.mmseqs2_cluster))
+all_prots       = df_all.shape[0]
+all_clusters    = len(set(df_all.mmseqs2_cluster))
+cov_prots       = df_cov.shape[0]
+cov_clusters    = len(set(df_cov.mmseqs2_cluster))
+allcov_prots    = df_allcov.shape[0]
+allcov_clusters = len(set(df_allcov.mmseqs2_cluster))
 print(
-    "All assemblies: %d proteins, %d clusters\nmetaT cov. >= 10: %d proteins, %d clusters" % (
-        df_all.shape[0],
-        len(set(df_all.mmseqs2_cluster)),
-        df_cov.shape[0],
-        len(set(df_cov.mmseqs2_cluster))
+"In total: %d proteins, %d clusters\n\
+Percentage values are computed w.r.t. total number of proteins/clusters:\n\
+1) Clusters representing ALL assemblies: %d proteins (%.2f%%), %d clusters (%.2f%%)\n\
+2) Clusters NOT repr. all assemblies AND metaT cov. >= 10: %d proteins (%.2f%%), %d clusters (%.2f%%)\n\
+3) Clusters repr. ALL assemblies AND metaT cov. >= 10: %d proteins (%.2f%%, %.2f%% of 1)), %d clusters (%.2f%%, %.2f%% of 1))\n" % (
+        total_prots, total_clusters,
+        all_prots, 100 * all_prots / total_prots, all_clusters, 100 * all_clusters / total_clusters,
+        cov_prots, 100 * cov_prots / total_prots, cov_clusters, 100 * cov_clusters / total_clusters,
+        allcov_prots, 100 * allcov_prots / total_prots, 100 * allcov_prots / all_prots, allcov_clusters, 100 * allcov_clusters / total_clusters, 100 * allcov_clusters / all_clusters
     )
 )
 
-# All assemblies: 428107 proteins, 51459 clusters
-# metaT cov. >= 10: 137054 proteins, 54747 clusters
+# In total: 1863851 proteins, 602303 clusters
+# Percentage values are computed w.r.t. total number of proteins/clusters:
+# 1) Clusters representing ALL assemblies: 428107 proteins (22.97%), 51459 clusters (8.54%)
+# 2) Clusters NOT repr. all assemblies AND metaT cov. >= 10: 137054 proteins (7.35%), 54747 clusters (9.09%)
+# 3) Clusters repr. ALL assemblies AND metaT cov. >= 10: 106660 proteins (5.72%, 24.91% of 1)), 13435 clusters (2.23%, 26.11% of 1))
 ```
\ No newline at end of file