Create a new helper function for generating k_values and GWA

2 files changed, 46 insertions, 0 deletions

diff --git a/gn3/computations/gemma.py b/gn3/computations/gemma.py
index ae04c40..5741692 100644
--- a/gn3/computations/gemma.py
+++ b/gn3/computations/gemma.py
@@ -5,8 +5,11 @@ import string
 
 from base64 import b64encode
 from hashlib import md5
+from typing import Dict
 from typing import List
+from typing import Optional
 from typing import ValuesView
+from gn3.commands import compose_gemma_cmd
 
 def generate_random_n_string(n_length: int) -> str:
     """Generate a random string that is N chars long"""
@@ -50,3 +53,26 @@ def do_paths_exist(paths: ValuesView) -> bool:
         if not os.path.isfile(path):
             return False
     return True
+
+
+def generate_gemma_computation_cmd(gemma_cmd: str,
+                                   gemma_kwargs: Dict[str, str],
+                                   output_file: str) -> Optional[str]:
+    """Create a higher order function that generates a command"""
+    geno_filename = gemma_kwargs.get("geno_filename", "")
+    trait_filename = gemma_kwargs.get("trait_filename")
+    ext, snps_filename = geno_filename.partition(".")[-1], ""
+    if geno_filename:
+        snps_filename = geno_filename.replace(f".{ext}", "")
+        snps_filename += f"_snps.{ext}"
+    _kwargs = {"g": geno_filename, "p": trait_filename}
+    if gemma_kwargs.get("covar_filename"):
+        _kwargs["a"] = gemma_kwargs.get("covar_filename")
+    if not do_paths_exist(_kwargs.values()):  # Prevents injection!
+        return None
+    if _kwargs.get("lmm"):
+        _kwargs["lmm"] = gemma_kwargs.get("lmm")
+    return compose_gemma_cmd(gemma_wrapper_cmd=gemma_cmd,
+                             gemma_kwargs=_kwargs,
+                             gemma_args=["-gk", ">",
+                                         output_file])
diff --git a/tests/unit/computations/test_gemma.py b/tests/unit/computations/test_gemma.py
index 23707ef..a7884ea 100644
--- a/tests/unit/computations/test_gemma.py
+++ b/tests/unit/computations/test_gemma.py
@@ -4,6 +4,7 @@ import unittest
 from unittest import mock
 from gn3.computations.gemma import generate_hash_of_string
 from gn3.computations.gemma import generate_pheno_txt_file
+from gn3.computations.gemma import generate_gemma_computation_cmd
 
 
 class TestGemma(unittest.TestCase):
@@ -29,3 +30,22 @@ class TestGemma(unittest.TestCase):
         """Test that a string is hashed correctly"""
         self.assertEqual(generate_hash_of_string("I^iQP&TlSR^z"),
                          "hMVRw8kbEp49rOmoIkhMjA")
+
+    @mock.patch("gn3.computations.gemma.do_paths_exist")
+    def test_compose_k_computation_cmd(self, mock_pathsp):
+        """Test that a K computation cmd is constructed properly"""
+        mock_pathsp.return_value = True
+        self.assertEqual(
+            generate_gemma_computation_cmd(
+                gemma_cmd="gemma-wrapper",
+                gemma_kwargs={
+                    "geno_filename": "genofile.txt",
+                    "trait_filename": "test.txt",
+                    "covar_filename": "genofile_snps.txt"},
+                output_file="/tmp/gn2/k_output_gUFhGu4rLG7k+CXLPk1OUg.txt",
+            ),
+            ("gemma-wrapper --json -- "
+             "-g genofile.txt -p "
+             "test.txt -a genofile_snps.txt "
+             "-gk > /tmp/gn2/"
+             "k_output_gUFhGu4rLG7k+CXLPk1OUg.txt"))