save_biasfactor_table now saves corrected coverage

Fixes a bug so that now when --save_biasfactor_table is specified, the corrected coverage is saved as well. Adds the --save_corrected_coverage_table flag/parameter.
epigen · May 4, 2022 · 1f90eb1 · 1f90eb1
1 parent 07ed4d7
commit 1f90eb1
Show file tree

Hide file tree

Showing 7 changed files with 44 additions and 12 deletions.
diff --git a/CHANGES.txt b/CHANGES.txt
@@ -1,2 +1,12 @@
 v0.5.0
-    Complete re-fractoring and overhaul.
+    Complete re-fractoring and overhaul.
+v0.5.1
+    Various minor bugfixes and improvements.
+v0.5.2
+    Improved ray import handling
+v0.5.3
+    Make ray-core an optional dependency in meta.yaml and provide details on this in docs
+v0.5.4
+    LIQUORICE_summary now handles samples with numerical names correctly
+v0.5.5
+    Fixes a bug so that now when --save_biasfactor_table is specified, the corrected coverage is saved as well. Adds the --save_corrected_coverage_table flag/parameter.
diff --git a/conda-recipe/meta.yaml b/conda-recipe/meta.yaml
@@ -1,9 +1,9 @@
 package:
   name: liquorice
-  version: 0.5.4
+  version: 0.5.5
 
 source:
-  url: https://github.com/epigen/LIQUORICE/archive/refs/tags/v0.5.4.tar.gz
+  url: https://github.com/epigen/LIQUORICE/archive/refs/tags/v0.5.5.tar.gz
   sha256: 476aae3c29516726cd4574b93ecb49bc89e85a931068bfba74a4d1b220361020
   #path: ..
 

diff --git a/dockerfile b/dockerfile
@@ -1,3 +1,3 @@
 FROM condaforge/mambaforge:4.9.2-5 as conda
 
-RUN mamba install -c bioconda -c conda-forge liquorice==0.5.4 ray-core
+RUN mamba install -c bioconda -c conda-forge liquorice==0.5.5 ray-core
diff --git a/docs/source/conf.py b/docs/source/conf.py
@@ -25,7 +25,7 @@
 author = 'Peter Peneder'
 
 # The full version, including alpha/beta/rc tags
-release = '0.5.4'
+release = '0.5.5'
 
 
 # -- General configuration ---------------------------------------------------

diff --git a/liquorice/LIQUORICE.py b/liquorice/LIQUORICE.py
@@ -34,7 +34,7 @@ def parse_args():
     :return: An `argparse.ArgumentParser` object storing the arguments.
     """
     parser = argparse.ArgumentParser(description="LIQUORICE: A tool for bias correction and quantification of changes "
-                                                 " in coverage around regions of interest in cfDNA WGS datasets",
+                                                 " in coverage around regions of interest in cfDNA WGS datasets. Documentation: https://liquorice.readthedocs.io; Publication: https://doi.org/10.1093/bioadv/vbac017.",
                                      formatter_class=argparse.ArgumentDefaultsHelpFormatter)
     required_keyword_args = parser.add_argument_group('Required named arguments')
 
@@ -186,9 +186,15 @@ def parse_args():
                                       "is not specified)",action="store_true")
 
     optional_keyword_args_output.add_argument(
-        '--save_biasfactor_table', help="If set, for each region-set, save a table of coverage and biasfactor values"
+        '--save_biasfactor_table', help="If set, for each region-set, save a table of bin coordinates, bin number, coverage, corrected coverage and biasfactor values"
                                         " per bin under "
-                                        "./<samplename>/<region-set name>/coverage_and_biasfactors_per_bin.csv",
+                                        "./<samplename>/<region-set name>/coverage_and_biasfactors_per_bin.csv. (Filesize can get quite large)",
+        action="store_true")
+
+    optional_keyword_args_output.add_argument(
+        '--save_corrected_coverage_table', help="If set, for each region-set, save a table of bin coordinates, bin number, coverage, and corrected coverage "
+                                        " per bin under "
+                                        "./<samplename>/<region-set name>/coverage_per_bin.csv",
         action="store_true")
 
     return parser
@@ -438,6 +444,7 @@ def main():
                 extend_to=args.extend_to,
                 n_cores=n_cpus_workflows,
                 save_biasfactor_table=args.save_biasfactor_table,
+                save_corrected_coverage_table=args.save_corrected_coverage_table,
                 use_default_fixed_sigma_values=True,
                 no_chr_prefix=args.no_chr_prefix,
                 percentile_split_core_rois=args.percentile_split_core_regions,
@@ -476,6 +483,7 @@ def main():
                 extend_to=args.extend_to,
                 n_cores=n_cpus_workflows,
                 save_biasfactor_table=args.save_biasfactor_table,
+                save_corrected_coverage_table=args.save_corrected_coverage_table,
                 use_default_fixed_sigma_values=True,
                 no_chr_prefix=args.no_chr_prefix,
                 save_training_table=args.save_training_table,

diff --git a/liquorice/utils/Workflows.py b/liquorice/utils/Workflows.py
@@ -226,6 +226,7 @@ def run_liquorice_on_regionset_with_pretrained_biasmodel(
     binsize: int =500, extend_to: int =20000,
     use_default_fixed_sigma_values: bool =True,
     save_biasfactor_table: bool =False,
+    save_corrected_coverage_table: bool =False,
     no_chr_prefix: bool=False,
     percentile_split_core_rois:bool=False,
     use_this_roi_biasfactortable:typing.Union[None,str]=None) -> None:
@@ -272,8 +273,10 @@ def run_liquorice_on_regionset_with_pretrained_biasmodel(
     :param use_default_fixed_sigma_values: If `True`, use the following contraints for the sigma values of the small,
         medium, and large Gaussian, repectively: 149-150 bp, 757-758 bp, and 6078-6079 bp.
     :param n_cores: Maximum number of cores to use during steps that allow multiprocessing/multithreading.
-    :param save_biasfactor_table: If `True`, save a table of coverage and biasfactor values per bin as
+    :param save_biasfactor_table: If `True`, save a table of bin coordinates, bin number, coverage, corrected coverage and biasfactor values under
         "coverage_and_biasfactors_per_bin.csv".
+    :param save_corrected_coverage_table: If `True`, save a table of bin coordinates, bin number, coverage, and corrected coverage under
+        "coverage_per_bin.csv".
     :param no_chr_prefix: If True, set the list of allowed chromosomes to [str(i) for i in range(1,23)] instead of
         ["chr"+str(i) for i in range(1,23)]
     :param percentile_split_core_rois: If set, split the central region into 5 bins of variable size instead of always
@@ -352,6 +355,10 @@ def run_liquorice_on_regionset_with_pretrained_biasmodel(
         logging.info(f"Writing bins and their bias factors to .csv")
         liq_table[[col for col in liq_table.columns if not col in ["sequence","mappability"]]].to_csv(
             "coverage_and_biasfactors_per_bin.csv",index=False)
+    if save_corrected_coverage_table:
+        logging.info(f"Writing bins and their bias factors to .csv")
+        liq_table[["chromosome","start","end","bin nr.","coverage","corrected coverage"]].to_csv(
+            "coverage_per_bin.csv",index=False)
 
     # Set up an object for plotting
     plotting = Plotting.Plotting(
@@ -452,6 +459,7 @@ def run_liquorice_train_biasmodel_on_same_regions(
         save_training_table: bool = False,
         use_default_fixed_sigma_values: bool =True,
         save_biasfactor_table: bool =False,
+        save_corrected_coverage_table: bool =False,
         no_chr_prefix: bool=False,
         percentile_split_core_rois:bool=False,
         use_cross_validated_predictions:bool=False,
@@ -501,8 +509,10 @@ def run_liquorice_train_biasmodel_on_same_regions(
     :param use_default_fixed_sigma_values: If `True`, use the following contraints for the sigma values of the small,
         medium, and large Gaussian, repectively: 149-150 bp, 757-758 bp, and 6078-6079 bp.
     :param n_cores: Maximum number of cores to use during steps that allow multiprocessing/multithreading.
-    :param save_biasfactor_table: If `True`, save a table of coverage and biasfactor values per bin as
+    :param save_biasfactor_table: If `True`, save a table of bin coordinates, bin number, coverage, corrected coverage and biasfactor values under
         "coverage_and_biasfactors_per_bin.csv".
+    :param save_corrected_coverage_table: If `True`, save a table of bin coordinates, bin number, coverage, and corrected coverage under
+        "coverage_per_bin.csv".
     :param no_chr_prefix: If True, set the list of allowed chromosomes to [str(i) for i in range(1,23)] instead of
         ["chr"+str(i) for i in range(1,23)]
     :param biasmodel_output_path: Path to which the trained biasmodel should be saved to. Must have a
@@ -616,8 +626,12 @@ def run_liquorice_train_biasmodel_on_same_regions(
     # write csv
     if save_biasfactor_table:
         logging.info(f"Writing bins and their bias factors to .csv")
-        liq_table_with_corr_cov[[col for col in liq_table.columns if not col in ["sequence","mappability"]]].to_csv(
+        liq_table_with_corr_cov[[col for col in liq_table_with_corr_cov.columns if not col in ["sequence","mappability"]]].to_csv(
             "coverage_and_biasfactors_per_bin.csv",index=False)
+    if save_corrected_coverage_table:
+        logging.info(f"Writing bins and their bias factors to .csv")
+        liq_table_with_corr_cov[["chromosome","start","end","bin nr.","coverage","corrected coverage"]].to_csv(
+            "coverage_per_bin.csv",index=False)
 
     # Set up an object for plotting
     plotting = Plotting.Plotting(

diff --git a/setup.cfg b/setup.cfg
@@ -1,6 +1,6 @@
 [metadata]
 name = liquorice
-version = 0.5.3
+version = 0.5.5
 author = Peter Peneder
 author_email = peter.peneder@ccri.at
 description = LIQUORICE: A tool for bias correction and quantification of changes in coverage around regions of interest in cfDNA WGS datasets