fix(pipeline.py): respect metadata locations

Previously, the call to construct the necessary dataframe had ignored any passed in metadata locations. Now, these locations rare referenced in construction, insuring they are available to all future methods.

lta/helpers/pipeline.py

@@ -76,11 +76,11 @@ def __post_init__(self) -> None:
                 self.file,
                 index_names=["Lipid", "Category", "m/z"],
                 column_names=[
-                    "Mode",
+                    self.mode,
                     "Sample",
-                    "Phenotype",
+                    self.level,
                     "Generation",
-                    "Tissue",
+                    self.tissue,
                     "Handling",
                 ],
                 index_col=[0, 1, 2],