From 41b4d11e66d663fc3327198751ee290a7b7c6c67 Mon Sep 17 00:00:00 2001
From: lala8 <lal.avantika@gene.com>
Date: Mon, 29 Jul 2024 17:07:34 +0000
Subject: [PATCH 1/2] allow custom blacklist file

---
 .github/workflows/pypi-publish.yml |  2 +-
 src/grelu/data/preprocess.py       | 14 ++++++++------
 tests/test_models.py               |  2 +-
 3 files changed, 10 insertions(+), 8 deletions(-)

diff --git a/.github/workflows/pypi-publish.yml b/.github/workflows/pypi-publish.yml
index 4223cca..30ffafb 100644
--- a/.github/workflows/pypi-publish.yml
+++ b/.github/workflows/pypi-publish.yml
@@ -5,7 +5,7 @@ name: Publish to PyPI
 
 on:
   push:
-    tags: 
+    tags:
       - "*"
 
 jobs:
diff --git a/src/grelu/data/preprocess.py b/src/grelu/data/preprocess.py
index b3d9a7d..ea25099 100644
--- a/src/grelu/data/preprocess.py
+++ b/src/grelu/data/preprocess.py
@@ -332,7 +332,7 @@ def filter_overlapping(
 
 def filter_blacklist(
     data: Union[pd.DataFrame, AnnData],
-    genome: str,
+    genome: Optional[str] = None,
     blacklist: Optional[str] = None,
     inplace: bool = False,
     window: int = 0,
@@ -344,7 +344,7 @@ def filter_blacklist(
         data: Either a pandas dataframe of genomic intervals or an Anndata
             object with intervals in .var
         genome: name of the genome corresponding to intervals
-        blacklist (str): path to blacklist file. If not given, will be
+        blacklist: path to blacklist file. If not given, it will be
             extracted from the package resources.
         inplace: If True, the input is modified in place. If False, a new
             dataframe or anndata object is returned.
@@ -357,11 +357,13 @@ def filter_blacklist(
     from grelu.resources import get_blacklist_file
 
     # Read blacklist
-    if genome is not None:
-        blacklist = get_blacklist_file(genome)
-
-    if isinstance(blacklist, str):
+    if genome is None:
+        assert (
+            blacklist is not None
+        ), "Either genome name or blacklist file must be provided"
         blacklist = read_bed(blacklist, str_index=False)
+    else:
+        blacklist = get_blacklist_file(genome)
 
     # Filter
     return filter_overlapping(
diff --git a/tests/test_models.py b/tests/test_models.py
index 2b4c546..93431a4 100644
--- a/tests/test_models.py
+++ b/tests/test_models.py
@@ -1,6 +1,6 @@
 import torch
-
 import wandb
+
 from grelu.model.models import (
     BorzoiModel,
     BorzoiPretrainedModel,

From 55ce1afa52ef3447222ed6176cbfc7a320ee26eb Mon Sep 17 00:00:00 2001
From: lala8 <lal.avantika@gene.com>
Date: Mon, 29 Jul 2024 17:29:15 +0000
Subject: [PATCH 2/2] fixed logic

---
 src/grelu/data/preprocess.py | 12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)

diff --git a/src/grelu/data/preprocess.py b/src/grelu/data/preprocess.py
index ea25099..2187bc8 100644
--- a/src/grelu/data/preprocess.py
+++ b/src/grelu/data/preprocess.py
@@ -356,14 +356,16 @@ def filter_blacklist(
     from grelu.io.bed import read_bed
     from grelu.resources import get_blacklist_file
 
-    # Read blacklist
-    if genome is None:
+    # Get path to blacklist file
+    if genome is not None:
+        blacklist = get_blacklist_file(genome)
+    else:
         assert (
             blacklist is not None
         ), "Either genome name or blacklist file must be provided"
-        blacklist = read_bed(blacklist, str_index=False)
-    else:
-        blacklist = get_blacklist_file(genome)
+
+    # Read blacklist file
+    blacklist = read_bed(blacklist, str_index=False)
 
     # Filter
     return filter_overlapping(