From d4fff5fda2d9fe2b9730a7cffcc8f85b3a8eff17 Mon Sep 17 00:00:00 2001
From: Frederick Muriuki Muriithi
Date: Thu, 28 Dec 2023 09:55:26 +0300
Subject: Handle transposed geno files.

---
 r_qtl/r_qtl2.py | 50 ++++++++++++++++++++++++++++++++++----------------
 1 file changed, 34 insertions(+), 16 deletions(-)

(limited to 'r_qtl')

diff --git a/r_qtl/r_qtl2.py b/r_qtl/r_qtl2.py
index e019d99..47f101e 100644
--- a/r_qtl/r_qtl2.py
+++ b/r_qtl/r_qtl2.py
@@ -32,7 +32,6 @@ def control_data(zfile: ZipFile) -> dict:
 
 def genotype_metadata(zfile: ZipFile, cdata: dict) -> dict:
     """Read Individual ID key and the marker names."""
-    # TODO: Handle transposed files
     line_num = 0
     with zfile.open(cdata["geno"]) as genofile:
         for line in filter(lambda line: not line.startswith("#"),
@@ -45,25 +44,48 @@ def genotype_metadata(zfile: ZipFile, cdata: dict) -> dict:
 
 def genotype_data(zfile: ZipFile, cdata: dict) -> Iterator[dict]:
     """Load the genotype file, making use of the control data."""
-    # TODO: Handle transposed files
-    with zfile.open(cdata["geno"]) as genofile:
-        reader = csv.DictReader(filter(lambda line: not line.startswith("#"),
-                                       io.TextIOWrapper(genofile)),
-                                delimiter=cdata.get("sep", ","))
-        if not cdata.get("geno_transposed", False):
+    def replace_genotype_codes(val):
+        return cdata["genotypes"].get(val, val)
+
+    def replace_na_strings(val):
+        return (None if val in cdata["na.strings"] else val)
+
+    if not cdata.get("geno_transposed", False):
+        with zfile.open(cdata["geno"]) as genofile:
+            reader = csv.DictReader(
+                filter(lambda line: not line.startswith("#"),
+                       io.TextIOWrapper(genofile)),
+                delimiter=cdata.get("sep", ","))
             for row in reader:
                 yield {
                     key: thread_op(
                         value,
-                        # replace genotype codes
-                        lambda val: cdata["genotypes"].get(val, val),
-                        # replace N/A strings
-                        lambda val: (
-                            None if val in cdata["na.strings"] else val))
+                        replace_genotype_codes,
+                        replace_na_strings)
                     for key,value
                     in row.items()
                 }
 
+    def __merge__(key, samples, line):
+        marker = line[0]
+        return (
+            dict(zip(
+                [key, marker],
+                (thread_op(item, replace_genotype_codes, replace_na_strings)
+                 for item in items)))
+            for items in zip(samples, line[1:]))
+
+    if cdata.get("geno_transposed", False):
+        with zfile.open(cdata["geno"]) as genofile:
+            lines = (line.strip().split(cdata.get("sep", ","))
+                     for line in filter(lambda line: not line.startswith("#"),
+                                         io.TextIOWrapper(genofile)))
+            id_line = next(lines)
+            id_key, samples = id_line[0], id_line[1:]
+            for line in lines:
+                for row in __merge__(id_key, samples, line):
+                    yield row
+
 def map_data(zfile: ZipFile, map_type: str, cdata: dict) -> dict:
     """Read gmap files to get the genome mapping data"""
     assert map_type in ("genetic-map", "physical-map"), "Invalid map type"
@@ -71,10 +93,6 @@ def map_data(zfile: ZipFile, map_type: str, cdata: dict) -> dict:
         "genetic-map": "gmap",
         "physical-map": "pmap"
     }[map_type]]
-    # TODO: Might need to check `gmap_transposed` and `pmap_transposed` instead
-    #  of `geno_transposed` -- see
-    #  https://github.com/rqtl/qtl2data/blob/main/ArabMAGIC/arabmagic_tair8.json
-    #  for the *_transposed values
     transposed_dict = {
         "genetic-map": "gmap_transposed",
         "physical-map": "pmap_transposed"
-- 
cgit v1.2.3