diff --git a/ablation/collate.py b/ablation/collate.py
index ade722865..3cd530bcd 100644
--- a/ablation/collate.py
+++ b/ablation/collate.py
@@ -1,5 +1,6 @@
 import logging
 import os
+from typing import Collection, Optional
 
 import click
 import pandas as pd
@@ -24,12 +25,13 @@ def read_collation() -> pd.DataFrame:
     return pykeen_report.utils.read_ablation_collation(COLLATION_PATH)
 
 
-def collate(key: str = 'hits@10') -> pd.DataFrame:
+def collate(key: str = 'hits@10', additional_metrics: Optional[Collection[str]] = None) -> pd.DataFrame:
     """Collate all results for a given metric."""
     return pykeen_report.utils.collate_ablation(
         results_directory=RESULTS,
         output_path=COLLATION_PATH,
         key=key,
+        additional_metrics=additional_metrics,
     )
 
 
diff --git a/src/pykeen_report/utils.py b/src/pykeen_report/utils.py
index 6bba6b056..f80ae43a0 100644
--- a/src/pykeen_report/utils.py
+++ b/src/pykeen_report/utils.py
@@ -7,7 +7,7 @@
 import os
 from copy import deepcopy
 from pathlib import Path
-from typing import Any, Iterable, Mapping, Optional, Type, Union
+from typing import Any, Collection, Iterable, Mapping, Optional, Type, Union
 
 import pandas as pd
 from tqdm import tqdm
@@ -33,11 +33,14 @@ def collate_ablation(
     results_directory: str,
     output_path: str,
     key: str,
+    additional_metrics: Optional[Collection[str]] = None,
 ) -> pd.DataFrame:
     """Collate all results for a given metric.
 
     :param key: The metric which you care about. Should be the same one against which you
      optimized
+    :param additional_metrics:
+        Additional metrics to collect.
     """
     columns = [
         'searcher',
@@ -55,6 +58,9 @@ def collate_ablation(
         'evaluation_time',
         key,
     ]
+    if additional_metrics is None:
+        additional_metrics = []
+    columns += list(additional_metrics)
 
     directories = [
         (directory, filenames)