Merge pull request #233 from AlexandrovLab/development

v1.3.5 Updates

Author: mdbarnesUCSD

CHANGELOG.md

@@ -6,6 +6,12 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 
 ## [Unreleased]
 
+## [1.3.5] - 2025-08-05
+
+### Changed
+- Updated README to clarify usage and examples.
+- Added custom output directory option.
+
 ## [1.3.4] - 2025-06-06
 
 ### Fixed

README.md

@@ -103,18 +103,18 @@ a DBS, SBS, ID, and TSB folder (there will also be a plots folder if this parame
 
 ***First six columns are required, and either the column "svclass" (deletion, translocation, tandem-duplication, or inversion) or the columns "strand1" & "strand2" (BRASS convention, see https://github.com/cancerit/BRASS/wiki/BEDPE) must also be present***
 
-
 ### Example with SV class present (tsv or csv file):
 
-| chrom1 | start1 | end1 | chrom2 | start2 | end2 | svclass |
-| :-----: | :-: | :-: | :-: | :-: | :-: | :-: |
-| 19 | 21268384 | 21268385 | 19 | 21327858 | 21327859 | deletion
+| sample  | chrom1 | start1   | end1     | chrom2 | start2   | end2     | svclass  |
+|:-------:|:------:|:--------:|:--------:|:------:|:--------:|:--------:|:--------:|
+| Sample1 | 19     | 21268384 | 21268385 | 19     | 21327858 | 21327859 | deletion |
 
 ### Example without SV class present (tsv or csv file):
 
-| chrom1 | start1 | end1 | chrom2 | start2 | end2 | strand1 | strand2
-| :-----: | :-: | :-: | :-: | :-: | :-: | :-: | :-: |
-| 19 | 21268384 | 21268385 | 19 | 21327858 | 21327859 | + | +
+| sample  | chrom1 | start1   | end1     | chrom2 | start2   | end2     | strand1 | strand2 |
+|:-------:|:------:|:--------:|:--------:|:------:|:--------:|:--------:|:-------:|:-------:|
+| Sample1 | 19     | 21268384 | 21268385 | 19     | 21327858 | 21327859 | +       | +       |
+
 
 ### Quick Start Example: ###
 

SigProfilerMatrixGenerator/controllers/cli_controller.py

@@ -143,6 +143,12 @@ def parse_arguments_matrix_generator(args: List[str]) -> argparse.Namespace:
         default=None,
     )
 
+    parser.add_argument(
+        "--output_directory",
+        help="Specify a custom output directory (default: None).",
+        default=None,
+    )
+
     result = parser.parse_args(args)
     return result
 
@@ -230,6 +236,7 @@ def dispatch_matrix_generator(self, user_args: List[str]) -> None:
             seqInfo=parsed_args.seqInfo,
             cushion=parsed_args.cushion,
             volume=parsed_args.volume,
+            output_directory=parsed_args.output_directory,
         )
 
     def dispatch_sv_matrix_generator(self, user_args: List[str]) -> None:

SigProfilerMatrixGenerator/scripts/SigProfilerMatrixGeneratorFunc.py

@@ -64,6 +64,7 @@ def SigProfilerMatrixGeneratorFunc(
     cushion=100,
     gs=False,
     volume=None,
+    output_directory=None,
 ):
     """
     Allows for the import of the sigProfilerMatrixGenerator.py function. Returns a dictionary
@@ -1188,37 +1189,40 @@ def SigProfilerMatrixGeneratorFunc(
             + " has not been installed\nPlease refer to the SigProfilerMatrixGenerator README for installation instructions:\n\thttps://github.com/AlexandrovLab/SigProfilerMatrixGenerator"
         )
 
-    # Organizes all of the input and output directories:
-    if path_to_input_files[-1] != "/":
-        path_to_input_files += "/"
-    vcf_path = path_to_input_files + "input/"
+    # Set base directory (for input/ and output/) based on user-provided output_directory if given
+    base_path = output_directory if output_directory else path_to_input_files
+    if not base_path.endswith("/"):
+        base_path += "/"
 
+    # Ensure input path and create if needed
+    vcf_path = os.path.join(base_path, "input/")
     vcf_path_original = vcf_path
+
+    # Copy all regular files from input source into vcf_path if vcf_path is missing or empty
     if not os.path.exists(vcf_path) or len(os.listdir(vcf_path)) < 1:
         os.makedirs(vcf_path, exist_ok=True)
-        input_files = os.listdir(path_to_input_files)
-        if os.path.exists(path_to_input_files + "input/"):
-            input_files.remove("input")
-        if os.path.exists(path_to_input_files + "logs/"):
-            input_files.remove("logs")
-        if ".DS_Store" in input_files:
-            input_files.remove(".DS_Store")
-        if "__init__.py" in input_files:
-            input_files.remove("__init__.py")
-        if "__pycache__" in input_files:
-            input_files.remove("__pycache__")
-        if os.path.exists(path_to_input_files + "output/"):
-            input_files.remove("output")
-        for files in input_files:
-            shutil.copy(path_to_input_files + files, vcf_path + files)
-    output_matrix = path_to_input_files + "output/"
-
-    if not os.path.exists(output_matrix):
-        os.makedirs(output_matrix)
+        if not path_to_input_files.endswith("/"):
+            path_to_input_files += "/"
+        for filename in os.listdir(path_to_input_files):
+            src = os.path.join(path_to_input_files, filename)
+            dst = os.path.join(vcf_path, filename)
+            if os.path.isfile(src):
+                shutil.copy(src, dst)
+
+    # Define the output matrix path, allowing override via `output_directory`
+    if output_directory:
+        if output_directory[-1] != "/":
+            output_directory += "/"
+        output_matrix = output_directory
+    else:
+        output_matrix = base_path + "output/"
+
+    os.makedirs(output_matrix, exist_ok=True)
+
 
     # Organizes the error and log files
     time_stamp = datetime.date.today()
-    output_log_path = path_to_input_files + "logs/"
+    output_log_path = output_matrix + "logs/"
     if not os.path.exists(output_log_path):
         os.makedirs(output_log_path)
     error_file = (

setup.py

@@ -3,7 +3,7 @@
 
 from setuptools import setup
 
-VERSION = "1.3.4"
+VERSION = "1.3.5"
 
 # remove the dist folder first if exists
 if os.path.exists("dist"):
@@ -23,7 +23,7 @@ def write_version_py(filename="SigProfilerMatrixGenerator/version.py"):
 # THIS FILE IS GENERATED FROM SIGPROFILEMATRIXGENERATOR SETUP.PY
 short_version = '%(version)s'
 version = '%(version)s'
-Update = 'v1.3.4: Fix mm39 and rn7 context counts and update in save_context_distribution script.'
+Update = 'v1.3.5: Update README and add custom output directory option'
 
 	"""
     fh = open(filename, "w")