From b4f2d314198ee9651f64f139ea1e8adf3e4250e1 Mon Sep 17 00:00:00 2001
From: Yasemin Bridges <y.bridges@qmul.ac.uk>
Date: Wed, 13 Nov 2024 13:51:49 +0000
Subject: [PATCH 1/2] allow specification of gene identifier

---
 src/phenotype2phenopacket/add/add_genes.py         |  9 ++++++---
 src/phenotype2phenopacket/cli_add.py               | 10 ++++++++++
 .../utils/phenopacket_utils.py                     |  4 ++--
 tests/test_phenopacket_utils.py                    | 14 +++++++-------
 4 files changed, 25 insertions(+), 12 deletions(-)

diff --git a/src/phenotype2phenopacket/add/add_genes.py b/src/phenotype2phenopacket/add/add_genes.py
index 2acf443..bbe14ba 100644
--- a/src/phenotype2phenopacket/add/add_genes.py
+++ b/src/phenotype2phenopacket/add/add_genes.py
@@ -1,7 +1,6 @@
 from pathlib import Path
 
 import polars as pl
-from phenopackets import Disease
 from pheval.utils.file_utils import all_files
 from pheval.utils.phenopacket_utils import (
     GeneIdentifierUpdater,
@@ -10,6 +9,7 @@
     phenopacket_reader,
 )
 
+from phenopackets import Disease
 from phenotype2phenopacket.utils.phenopacket_utils import (
     PhenopacketInterpretationExtender,
     PhenopacketUtil,
@@ -68,19 +68,22 @@ def add_genes(
         )
 
 
-def add_genes_to_directory(phenopacket_dir: Path, genes_to_disease: pl.DataFrame, output_dir: Path):
+def add_genes_to_directory(
+    phenopacket_dir: Path, genes_to_disease: pl.DataFrame, gene_identifier: str, output_dir: Path
+):
     """
     Add known gene-to-phenotype relationships to the interpretations of a directory of phenopackets.
 
     Args:
         phenopacket_dir (Path): Directory containing the phenopacket files.
         genes_to_disease (pl.DataFrame): DataFrame containing genes_to_disease.txt entries.
+        gene_identifier (str): Gene identifier for the phenopacket.
         output_dir (Path): Directory to store the updated phenopackets.
     """
     hgnc_dict = create_hgnc_dict()
     identifier_map = create_gene_identifier_map()
     gene_identifier_updater = GeneIdentifierUpdater(
-        gene_identifier="ensembl_id", hgnc_data=hgnc_dict, identifier_map=identifier_map
+        gene_identifier=gene_identifier, hgnc_data=hgnc_dict, identifier_map=identifier_map
     )
     for phenopacket_path in all_files(phenopacket_dir):
         add_genes(phenopacket_path, genes_to_disease, gene_identifier_updater, output_dir)
diff --git a/src/phenotype2phenopacket/cli_add.py b/src/phenotype2phenopacket/cli_add.py
index af642ca..7fc1fbf 100644
--- a/src/phenotype2phenopacket/cli_add.py
+++ b/src/phenotype2phenopacket/cli_add.py
@@ -28,9 +28,17 @@
     help="Path to output directory.",
     type=Path,
 )
+@click.option(
+    "--gene-identifier",
+    "-i",
+    required=False,
+    help="Gene identifier to update in phenopacket",
+    type=click.Choice(["ensembl_id", "entrez_id", "hgnc_id"]),
+)
 def add_genes_command(
     phenopacket_dir: Path,
     genes_to_disease: Path,
+    gene_identifier: str,
     output_dir: Path,
 ):
     """
@@ -39,6 +47,7 @@ def add_genes_command(
     Args:
         phenopacket_dir (Path): Directory containing the phenopacket files.
         genes_to_disease (Path): Path to the genes_to_disease.txt file.
+        gene_identifier (str): Gene identifier to add in phenopacket
         output_dir (Path): Directory to store the updated phenopackets.
     """
     output_dir.mkdir(exist_ok=True)
@@ -46,5 +55,6 @@ def add_genes_command(
     add_genes_to_directory(
         phenopacket_dir,
         genes_to_disease_df,
+        gene_identifier,
         output_dir,
     )
diff --git a/src/phenotype2phenopacket/utils/phenopacket_utils.py b/src/phenotype2phenopacket/utils/phenopacket_utils.py
index 540e258..147b0b4 100644
--- a/src/phenotype2phenopacket/utils/phenopacket_utils.py
+++ b/src/phenotype2phenopacket/utils/phenopacket_utils.py
@@ -11,6 +11,8 @@
 import polars as pl
 from google.protobuf.timestamp_pb2 import Timestamp
 from oaklib.implementations import ProntoImplementation
+from pheval.utils.phenopacket_utils import GeneIdentifierUpdater, create_json_message
+
 from phenopackets import (
     Age,
     Diagnosis,
@@ -26,8 +28,6 @@
     Resource,
     TimeElement,
 )
-from pheval.utils.phenopacket_utils import GeneIdentifierUpdater, create_json_message
-
 from phenotype2phenopacket.utils.utils import is_float
 
 
diff --git a/tests/test_phenopacket_utils.py b/tests/test_phenopacket_utils.py
index 9a2059c..7534ae1 100644
--- a/tests/test_phenopacket_utils.py
+++ b/tests/test_phenopacket_utils.py
@@ -3,6 +3,13 @@
 from unittest.mock import Mock, patch
 
 import polars as pl
+from pheval.utils.phenopacket_utils import (
+    GeneIdentifierUpdater,
+    create_gene_identifier_map,
+    create_hgnc_dict,
+)
+from polars.testing import assert_frame_equal
+
 from phenopackets import (
     Age,
     Diagnosis,
@@ -19,13 +26,6 @@
     Resource,
     TimeElement,
 )
-from pheval.utils.phenopacket_utils import (
-    GeneIdentifierUpdater,
-    create_gene_identifier_map,
-    create_hgnc_dict,
-)
-from polars.testing import assert_frame_equal
-
 from phenotype2phenopacket.utils.phenopacket_utils import (
     OnsetTerm,
     PhenopacketInterpretationExtender,

From 1e96ca3e0b8f4de35d8e5e6f002dd6ba91d16cdb Mon Sep 17 00:00:00 2001
From: Yasemin Bridges <y.bridges@qmul.ac.uk>
Date: Wed, 13 Nov 2024 13:52:04 +0000
Subject: [PATCH 2/2] bump version

---
 pyproject.toml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/pyproject.toml b/pyproject.toml
index 873201e..ad4e214 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "phenotype2phenopacket"
-version = "0.6.3"
+version = "0.6.4"
 description = ""
 authors = ["Yasemin Bridges <y.bridges@qmul.ac.uk>"]
 readme = "README.md"