From aff82f65eca0a4224a36bb647598ba1c4915ecd5 Mon Sep 17 00:00:00 2001
From: Aaron Mussig <aaronmussig@gmail.com>
Date: Wed, 9 Mar 2022 17:09:01 +1000
Subject: [PATCH] feat(GUNC): Added GTDB R95 dataset methods.

---
 magna/gunc/__init__.py | 26 ++++++++++++++++++++++++++
 setup.py               |  2 +-
 2 files changed, 27 insertions(+), 1 deletion(-)
 create mode 100644 magna/gunc/__init__.py

diff --git a/magna/gunc/__init__.py b/magna/gunc/__init__.py
new file mode 100644
index 0000000..b9e014d
--- /dev/null
+++ b/magna/gunc/__init__.py
@@ -0,0 +1,26 @@
+import os
+
+import pandas as pd
+
+from magna.config import MAGNA_DIR
+
+
+def gunc_max_css_scores_gtdb_r95() -> pd.DataFrame:
+    path = os.path.join(MAGNA_DIR, 'dataset', 'gunc', 'GUNC.gtdb_95.maxCSS_level.feather')
+    if not os.path.isfile(path):
+        raise IOError(f'{path} does not exist.')
+    return pd.read_feather(path)
+
+
+def gunc_contig_assignment_gtdb_r95() -> pd.DataFrame:
+    path = os.path.join(MAGNA_DIR, 'dataset', 'gunc', 'GUNC.gtdb_95.contig_assignments.feather')
+    if not os.path.isfile(path):
+        raise IOError(f'{path} does not exist.')
+    return pd.read_feather(path)
+
+
+def gunc_all_levels_gtdb_r95() -> pd.DataFrame:
+    path = os.path.join(MAGNA_DIR, 'dataset', 'gunc', 'gtdb_95.all_levels.tsv')
+    if not os.path.isfile(path):
+        raise IOError(f'{path} does not exist.')
+    return pd.read_feather(path)
diff --git a/setup.py b/setup.py
index 56feacd..9e231dd 100644
--- a/setup.py
+++ b/setup.py
@@ -43,6 +43,6 @@ def readme():
       ],
       packages=find_packages(),
       include_package_data=True,
-      install_requires=['tqdm', 'pandas', 'pyarrow', 'numpy', 'dendropy'],
+      install_requires=['tqdm', 'pandas>=1.1.0', 'pyarrow', 'numpy', 'dendropy', 'biopython'],
       python_requires='>=3.6',
       )