fix: Make genome assembly optional. Use builtin chromsizes by default (#20)

Author: nvictus

CHANGELOG.md

@@ -6,6 +6,13 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 
 ---
 
+## [1.0.5] - December 29, 2025
+
+### Updated
+- "assembly" argument is optional. Assembly is inferred from Cooler if possible, with fallback to optional argument.
+
+---
+
 ## [1.0.4] - October 31, 2025
 
 ### Added

jointly_hic/__init__.py

@@ -1 +1 @@
-__version__ = "1.0.4"
+__version__ = "1.0.5"

jointly_hic/core/decomposer.py

@@ -146,8 +146,13 @@ def partition(self) -> np.ndarray:
         return self._partition
 
     def get_chromosome_sizes(self) -> pd.Series:
-        """Download chromosome sizes using bioframe."""
-        chromosome_sizes = bioframe.fetch_chromsizes(self.configuration.assembly.lower())
+        """Obtain chromosome sizes from cooler or using bioframe."""
+        if self.configuration.assembly.lower() == "unknown":
+            chromosome_sizes = Cooler(
+                f"{self.configuration.mcools[0]}::/resolutions/{self.configuration.resolution}"
+            ).chromsizes
+        else:
+            chromosome_sizes = bioframe.fetch_chromsizes(self.configuration.assembly.lower())
         # chromosome_sizes is a pandas Series with chromosome names as index and chromosome sizes as values
         # Filter to only the first chrom_limit chromosomes
         chromosome_sizes = chromosome_sizes.iloc[: self.configuration.chrom_limit]

jointly_hic/parser.py

@@ -85,12 +85,6 @@ def configure_embed_parser(self, embed_parser):
             required=True,
             type=int,
         )
-        embed_parser.add_argument(
-            "--assembly",
-            help="Genome assembly name (UCSC or NCBI) used for alignment (e.g. hg38, mm10, ce11)",
-            required=True,
-            type=str,
-        )
         embed_parser.add_argument(
             "--output",
             help="Prefix for output files (Default: output_<datetime>)",
@@ -149,6 +143,13 @@ def configure_embed_parser(self, embed_parser):
             type=int,
             default=10000,
         )
+        embed_parser.add_argument(
+            "--assembly",
+            help="Optional genome assembly name (UCSC or NCBI) used for alignment (e.g. hg38, mm10, ce11)",
+            required=False,
+            type=str,
+            default="unknown",
+        )
         embed_parser.set_defaults(func=self.run_embedding)
 
     def configure_post_process_parser(self, post_process_parser):