Update paths for consensus peaks and bigwigs

seqnado/utils.py

@@ -683,7 +683,10 @@ def from_dataframe(cls, df: pd.DataFrame, simplified: bool = True, **kwargs):
 
         return cls(assays=experiments, **kwargs)
 
-def symlink_file(output_dir: pathlib.Path, source_path: pathlib.Path, new_file_name: str):
+
+def symlink_file(
+    output_dir: pathlib.Path, source_path: pathlib.Path, new_file_name: str
+):
     """
     Create a symlink in the output directory with the new file name.
     """
@@ -694,13 +697,16 @@ def symlink_file(output_dir: pathlib.Path, source_path: pathlib.Path, new_file_n
         except FileExistsError:
             logger.warning(f"Symlink for {new_path} already exists.")
 
-def symlink_fastq_files(design: Union[Design, DesignIP], output_dir: str = "seqnado_output/fastqs/") -> None:
+
+def symlink_fastq_files(
+    design: Union[Design, DesignIP], output_dir: str = "seqnado_output/fastqs/"
+) -> None:
     """
     Symlink the fastq files to the output directory.
     """
     output_dir = pathlib.Path(output_dir)
     output_dir.mkdir(parents=True, exist_ok=True)
-    
+
     if isinstance(design, Design):
         for assay_name, assay in design.assays.items():
             symlink_file(output_dir, assay.r1.path, f"{assay_name}_1.fastq.gz")
@@ -713,14 +719,16 @@ def symlink_fastq_files(design: Union[Design, DesignIP], output_dir: str = "seqn
             ip_assay = experiment.ip_files
             symlink_file(output_dir, ip_assay.r1.path, f"{ip_assay.name}_1.fastq.gz")
             if ip_assay.is_paired:
-                symlink_file(output_dir, ip_assay.r2.path, f"{ip_assay.name}_2.fastq.gz")
+                symlink_file(
+                    output_dir, ip_assay.r2.path, f"{ip_assay.name}_2.fastq.gz"
+                )
 
             if experiment.control_files:
                 control_assay = experiment.control_files
-                control_r1_name = control_assay.r1.path.name 
+                control_r1_name = control_assay.r1.path.name
                 symlink_file(output_dir, control_assay.r1.path, control_r1_name)
                 if control_assay.is_paired:
-                    control_r2_name = control_assay.r2.path.name 
+                    control_r2_name = control_assay.r2.path.name
                     symlink_file(output_dir, control_assay.r2.path, control_r2_name)
 
 
@@ -810,7 +818,7 @@ def define_output_files(
             for group_name, df in snakemake_design.to_dataframe().groupby("merge"):
                 assay_output.extend(
                     expand(
-                        "seqnado_output/consensus_peaks/{group_name}.bed",
+                        "seqnado_output/peaks/consensus/{group_name}.bed",
                         group_name=group_name,
                     )
                 )

seqnado/workflow/rules/consensus_peaks.smk

@@ -11,7 +11,7 @@ rule merge_bams:
     input:
         bams=get_bam_files_for_merge,
     output:
-        "seqnado_output/consensus_peaks/{group}.bam",
+        "seqnado_output/consensus_peaks/bam/{group}.bam",
     threads: 8
     log:
         "seqnado_output/consensus_peaks/{group}.log",
@@ -23,28 +23,30 @@ rule merge_bams:
 
 use rule index_bam as index_consensus_bam with:
     input:
-        bam="seqnado_output/consensus_peaks/{sample}.bam",
+        bam="seqnado_output/consensus_peaks/bam/{group}.bam",
     output:
-        bai="seqnado_output/consensus_peaks/{sample}.bam.bai",
+        bai="seqnado_output/consensus_peaks/bam/{group}.bam.bai",
     threads: 8
 
 
 use rule deeptools_make_bigwigs as deeptools_make_bigwigs_consensus with:
     input:
-        bam="seqnado_output/consensus_peaks/{sample}.bam",
-        bai="seqnado_output/consensus_peaks/{sample}.bam.bai",
+        bam="seqnado_output/consensus_peaks/bam/{group}.bam",
+        bai="seqnado_output/consensus_peaks/bam/{group}.bam.bai",
     output:
-        bigwig="seqnado_output/consensus_peaks/{sample}.bigWig",
+        bigwig="seqnado_output/bigwigs/consensus/{group}.bigWig",
     threads: 8
+    log:
+        "seqnado_output/logs/consensus_peaks/bigwigs/{group}.log",
 
 
 rule lanceotron_no_input_consensus:
     input:
-        group="seqnado_output/consensus_peaks/{group}.bigWig",
+        group="seqnado_output/bigwigs/consensus/{group}.bigWig",
     output:
-        peaks="seqnado_output/consensus_peaks/{group}.bed",
+        peaks="seqnado_output/peaks/consensus/{group}.bed",
     log:
-        "seqnado_output/consensus_peaks/{group}.log",
+        "seqnado_output/logs/consensus_peaks/peaks/{group}.log",
     params:
         options=seqnado.utils.check_options(config["lanceotron"]["callpeak"]),
         outdir=lambda wc, output: os.path.dirname(output.peaks),
@@ -63,9 +65,9 @@ rule lanceotron_no_input_consensus:
 
 use rule bed_to_bigbed as bed_to_bigbed_consensus with:
     input:
-        bed="seqnado_output/consensus_peaks/{sample}.bed",
+        bed="seqnado_output/peaks/consensus/{group}.bed",
     output:
-        bigbed="seqnado_output/consensus_peaks/{sample}.bigBed",
+        bigbed="seqnado_output/peaks/consensus/{group}.bigBed",
     threads: 1
     log:
-        "seqnado_output/consensus_peaks/{sample}.log",
+        "seqnado_output/logs/consensus_peaks/peaks/{group}.log",

seqnado/workflow/rules/hub.smk

@@ -93,7 +93,7 @@ def get_hub_input(wildcards):
     if "merge" in DESIGN.to_dataframe().columns:
         input_files.extend(
             expand(
-                "seqnado_output/consensus_peaks/{group}.bigWig",
+                "seqnado_output/bigwigs/consensus/{group}.bigWig",
                 group=DESIGN.to_dataframe()["merge"].unique(),
             )
         )
@@ -125,7 +125,7 @@ def get_peak_files(wildcards):
     if "merge" in DESIGN.to_dataframe().columns:
         peak_files.extend(
             expand(
-                "seqnado_output/consensus_peaks/{group}.bigBed",
+                "seqnado_output/consensus_peaks/peaks/{group}.bigBed",
                 group=DESIGN.to_dataframe().merge.unique(),
             )
         )
@@ -207,6 +207,7 @@ rule generate_hub:
     params:
         **get_hub_params(config),
         assay=ASSAY,
+        has_consensus_peaks="merge" in DESIGN.to_dataframe().columns,
     script:
         "../scripts/create_hub.py"
 

seqnado/workflow/scripts/create_hub.py

@@ -7,6 +7,7 @@
 from loguru import logger
 import tracknado
 
+
 def get_rna_samplename(path: str):
     p = pathlib.Path(path)
     return re.split(r"_[plus|minus]", p.name)[0]
@@ -17,6 +18,12 @@ def get_rna_samplename(path: str):
     snakemake.input.data,
     columns=["fn"],
 )
+
+
+# Use the TrackFiles class to deduplicate files and add metadata
+df = tracknado.TrackFiles(files=df, deduplicate=True).files
+
+
 if snakemake.params.assay == "ChIP":
     df[["samplename", "antibody"]] = df["fn"].str.extract(
         r".*/(.*)_(.*)\.(?:bigBed|bigWig)"
@@ -32,14 +39,13 @@ def get_rna_samplename(path: str):
     df["method"] = df["fn"].apply(lambda x: x.split("/")[-2])
     df["strand"] = np.where(df["fn"].str.contains("_plus.bigWig"), "plus", "minus")
 
-# remove dupilcates from design
-df.drop_duplicates()
-
 # Create hub design
 design = tracknado.TrackDesign.from_design(
     df,
     color_by=snakemake.params.color_by,
-    subgroup_by=snakemake.params.subgroup_by,
+    subgroup_by=snakemake.params.subgroup_by
+    if any(snakemake.params.subgroup_by)
+    else None,
     supergroup_by=snakemake.params.supergroup_by,
     overlay_by=snakemake.params.overlay_by,
 )