Fix problematic psirc output

conf/modules.config

@@ -697,7 +697,7 @@ process {
             [
                 "-v FS='\\t'",
                 "-v OFS='\\t'",
-                "'{ \$4 = \$1 \":\" \$2 \"-\" \$3",
+                "'{ \$4 = \"circ_\" \$1 \":\" \$2 \"-\" \$3",
                 (params.consider_strand ? " \":\" \$6" : ""),
                 "; print }'",
             ].join(' ').trim()

modules/local/annotation/bed2gtf/environment.yml

@@ -2,4 +2,5 @@ channels:
   - conda-forge
   - bioconda
 dependencies:
-  - conda-forge::polars=1.24.0
+  - conda-forge::polars=1.31.0
+  - conda-forge::pyyaml=6.0.2

modules/local/annotation/bed2gtf/main.nf

@@ -4,8 +4,8 @@ process ANNOTATION_BED2GTF {
 
     conda "${moduleDir}/environment.yml"
     container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
-        'oras://community.wave.seqera.io/library/polars:1.24.0--800cd3e4ff805434' :
-        'community.wave.seqera.io/library/polars:1.24.0--2d2d323e8514e707' }"
+        'https://community-cr-prod.seqera.io/docker/registry/v2/blobs/sha256/78/785e4d9624ef7fb24cfee74f815d3e69540e99e0ae299cbd333e047e08706f7e/data' :
+        'community.wave.seqera.io/library/polars_pyyaml:e53e9c9a38a99374' }"
 
     input:
     tuple val(meta), path(bed12), path(db_intersections)

modules/local/annotation/bed2gtf/templates/bed2gtf.py

@@ -1,27 +1,23 @@
 #!/usr/bin/env python
 
 import platform
+import yaml
 
 import polars as pl
 
-def format_yaml_like(data: dict, indent: int = 0) -> str:
-    """Formats a dictionary to a YAML-like string.
+# Versions
+
+versions = {
+    "${task.process}": {
+        "python": platform.python_version(),
+        "polars": pl.__version__,
+    }
+}
 
-    Args:
-        data (dict): The dictionary to format.
-        indent (int): The current indentation level.
+with open("versions.yml", "w") as f:
+    f.write(yaml.dump(versions))
 
-    Returns:
-        str: A string formatted as YAML.
-    """
-    yaml_str = ""
-    for key, value in data.items():
-        spaces = "  " * indent
-        if isinstance(value, dict):
-            yaml_str += f"{spaces}{key}:\\n{format_yaml_like(value, indent + 1)}"
-        else:
-            yaml_str += f"{spaces}{key}: {value}\\n"
-    return yaml_str
+# Main
 
 exons_only = bool("${exons_only}")
 
@@ -31,7 +27,12 @@ def format_yaml_like(data: dict, indent: int = 0) -> str:
             'readNumber', 'circType', 'gene', 'transcript',
             'index', 'flankIntron'
             ]
-df = pl.scan_csv('${bed12}', separator='\\t', has_header=False, new_columns=columns)
+try:
+    df = pl.scan_csv('${bed12}', separator='\\t', has_header=False, new_columns=columns, raise_if_empty=True)
+except pl.exceptions.NoDataError:
+    with open('${prefix}.${suffix}', 'w') as f:
+        f.write('')
+    exit(0)
 
 df = df.with_columns(
     attributes = pl.lit('gene_id "') + pl.col('gene') + pl.lit('"; transcript_id "') + pl.col('name') + pl.lit('";'),
@@ -77,15 +78,3 @@ def format_yaml_like(data: dict, indent: int = 0) -> str:
 df_combined = df_combined.sort('chr', 'start', 'end')
 
 df_combined.collect().write_csv('${prefix}.${suffix}', separator='\\t', include_header=False, quote_style="never")
-
-# Versions
-
-versions = {
-    "${task.process}": {
-        "python": platform.python_version(),
-        "polars": pl.__version__,
-    }
-}
-
-with open("versions.yml", "w") as f:
-    f.write(format_yaml_like(versions))

modules/local/combinebeds/counts/environment.yml

@@ -2,5 +2,5 @@ channels:
   - conda-forge
   - bioconda
 dependencies:
-  - conda-forge::polars=1.8.2
-  - conda-forge::upsetplot=0.9.0
+  - conda-forge::polars=1.31.0
+  - conda-forge::pyyaml=6.0.2

modules/local/combinebeds/counts/main.nf

@@ -4,8 +4,8 @@ process COMBINEBEDS_COUNTS {
 
     conda "${moduleDir}/environment.yml"
     container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
-        'oras://community.wave.seqera.io/library/pandas_polars_pyarrow_upsetplot:8840b96e156438fc' :
-        'community.wave.seqera.io/library/pandas_polars_pyarrow_upsetplot:6982d93f61d3e2ff' }"
+        'https://community-cr-prod.seqera.io/docker/registry/v2/blobs/sha256/78/785e4d9624ef7fb24cfee74f815d3e69540e99e0ae299cbd333e047e08706f7e/data' :
+        'community.wave.seqera.io/library/polars_pyyaml:e53e9c9a38a99374' }"
 
     input:
     tuple val(meta), val(aggregation), path(candidates), path(beds)

modules/local/combinebeds/counts/templates/counts.py

@@ -1,29 +1,23 @@
 #!/usr/bin/env python
 
 import platform
-import base64
-import json
+import yaml
 
 import polars as pl
 
-def format_yaml_like(data: dict, indent: int = 0) -> str:
-    """Formats a dictionary to a YAML-like string.
-
-    Args:
-        data (dict): The dictionary to format.
-        indent (int): The current indentation level.
-
-    Returns:
-        str: A string formatted as YAML.
-    """
-    yaml_str = ""
-    for key, value in data.items():
-        spaces = "  " * indent
-        if isinstance(value, dict):
-            yaml_str += f"{spaces}{key}:\\n{format_yaml_like(value, indent + 1)}"
-        else:
-            yaml_str += f"{spaces}{key}: {value}\\n"
-    return yaml_str
+# Versions
+
+versions = {
+    "${task.process}": {
+        "python": platform.python_version(),
+        "polars": pl.__version__
+    }
+}
+
+with open("versions.yml", "w") as f:
+    f.write(yaml.dump(versions))
+
+# Main
 
 max_shift = int("${max_shift}")
 consider_strand = "${consider_strand}" == "true"
@@ -37,11 +31,25 @@ def format_yaml_like(data: dict, indent: int = 0) -> str:
 
 columns = ["chr", "start", "end", "name", "score", "strand"]
 
-df_candidates = pl.scan_csv(candidate_path, has_header=False, separator="\\t", new_columns=columns)
+try:
+    df_candidates = pl.scan_csv(candidate_path, has_header=False, separator="\\t", new_columns=columns, raise_if_empty=True)
+except pl.exceptions.NoDataError:
+    print("No data in ${candidates}")
+    with open(f"{prefix}.{suffix}", "w") as f:
+        f.write('')
+    exit(0)
+
 df_candidates = df_candidates.select(columns)
 df_candidates = df_candidates.with_columns(sample=pl.lit("candidate"), tool=pl.lit("candidate"), score=pl.lit(None))
 
-df = pl.scan_csv(bed_paths, has_header=False, separator="\\t", new_columns=columns + ["sample", "tool"])
+try:
+    df = pl.scan_csv(bed_paths, has_header=False, separator="\\t", new_columns=columns + ["sample", "tool"], raise_if_empty=True)
+except pl.exceptions.NoDataError:
+    print("No data in ${beds}")
+    with open(f"{prefix}.{suffix}", "w") as f:
+        f.write('')
+    exit(0)
+
 df_combined = pl.concat([df, df_candidates])
 
 df_combined = df_combined.sort("end"  ).with_columns(end_group  =pl.col("end"  ).diff().fill_null(0).gt(max_shift).cum_sum())
@@ -62,7 +70,14 @@ def format_yaml_like(data: dict, indent: int = 0) -> str:
 df = df.filter((pl.col("start") - pl.col("start_right")).abs() <= max_shift)
 df = df.filter((pl.col("end") - pl.col("end_right")).abs() <= max_shift)
 df = df.group_by(["chr", "start", "end", "strand", "start_group", "end_group", "sample", "tool"]).agg(score=pl.sum("score"))
-df = df.collect().lazy()
+
+try:
+    df = df.collect().lazy()
+except pl.exceptions.NoDataError:
+    print("No data after processing")
+    with open(f"{prefix}.{suffix}", "w") as f:
+        f.write('')
+    exit(0)
 
 samples = df.select("sample").group_by("sample").len().collect()["sample"].to_list()
 df = df.collect().pivot(on="sample", values="score", index=["chr", "start", "end", "strand", "start_group", "end_group"], aggregate_function=aggregation).lazy()
@@ -73,15 +88,3 @@ def format_yaml_like(data: dict, indent: int = 0) -> str:
 df = df.fill_null(0)
 
 df.sink_csv(f"{prefix}.{suffix}", separator="\\t", include_header=True)
-
-# Versions
-
-versions = {
-    "${task.process}": {
-        "python": platform.python_version(),
-        "polars": pl.__version__
-    }
-}
-
-with open("versions.yml", "w") as f:
-    f.write(format_yaml_like(versions))

modules/local/combinebeds/shifts/environment.yml

@@ -2,6 +2,7 @@ channels:
   - conda-forge
   - bioconda
 dependencies:
-  - conda-forge::polars=1.8.2
   - conda-forge::altair=5.5.0
-  - conda-forge::vl-convert-python==1.7.0
+  - conda-forge::polars=1.31.0
+  - conda-forge::pyyaml=6.0.2
+  - conda-forge::vl-convert-python=1.7.0

modules/local/combinebeds/shifts/main.nf

@@ -4,15 +4,15 @@ process COMBINEBEDS_SHIFTS {
 
     conda "${moduleDir}/environment.yml"
     container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
-        'oras://community.wave.seqera.io/library/altair_polars_vl-convert-python:e6f1dca28de76d13' :
-        'community.wave.seqera.io/library/altair_polars_vl-convert-python:a6c5ee679445250d' }"
+        'https://community-cr-prod.seqera.io/docker/registry/v2/blobs/sha256/cb/cb4ec5e7ad6feeda9b1cd6d194043b6cc0d2952c93a28cddc98fc7d67c078141/data' :
+        'community.wave.seqera.io/library/altair_polars_pyyaml_vl-convert-python:c19053ed9a1a6146' }"
 
     input:
     tuple val(meta), path(beds)
 
     output:
-    path "*.png"       , emit: plots
-    path "*.json"      , emit: multiqc
+    path "*.png"       , emit: plots, optional: true
+    path "*.json"      , emit: multiqc, optional: true
     path "versions.yml", emit: versions
 
     script:

modules/local/combinebeds/shifts/templates/shifts.py

@@ -4,35 +4,36 @@
 import base64
 import json
 from itertools import product
+import yaml
 
 import polars as pl
 import altair as alt
 
-def format_yaml_like(data: dict, indent: int = 0) -> str:
-    """Formats a dictionary to a YAML-like string.
-
-    Args:
-        data (dict): The dictionary to format.
-        indent (int): The current indentation level.
-
-    Returns:
-        str: A string formatted as YAML.
-    """
-    yaml_str = ""
-    for key, value in data.items():
-        spaces = "  " * indent
-        if isinstance(value, dict):
-            yaml_str += f"{spaces}{key}:\\n{format_yaml_like(value, indent + 1)}"
-        else:
-            yaml_str += f"{spaces}{key}: {value}\\n"
-    return yaml_str
+# Versions
+
+versions = {
+    "${task.process}": {
+        "python": platform.python_version(),
+        "polars": pl.__version__,
+        "altair": alt.__version__
+    }
+}
+
+with open("versions.yml", "w") as f:
+    f.write(yaml.dump(versions))
+
+# Main
 
 meta_id = "${meta.id}"
 
-df = pl.scan_csv("${beds}".split(" "),
-                    separator="\\t",
-                    has_header=False,
-                    new_columns=["chr", "start", "end", "name", "score", "strand", "sample", "tool"])
+try:
+    df = pl.scan_csv("${beds}".split(" "),
+                        separator="\\t",
+                        has_header=False,
+                        raise_if_empty=True,
+                        new_columns=["chr", "start", "end", "name", "score", "strand", "sample", "tool"])
+except pl.exceptions.NoDataError:
+    exit(0)
 
 df = df.group_by("chr", "start", "end", "strand").agg(tools=pl.col("tool").unique(), samples=pl.col("sample").unique())
 
@@ -105,16 +106,3 @@ def get_group_sizes(df: pl.LazyFrame, max_shift: int, consider_strand: bool) ->
 
     with open(f"{metric}.shifts_mqc.json", "w") as f:
         f.write(json.dumps(multiqc, indent=4))
-
-# Versions
-
-versions = {
-    "${task.process}": {
-        "python": platform.python_version(),
-        "polars": pl.__version__,
-        "altair": alt.__version__
-    }
-}
-
-with open("versions.yml", "w") as f:
-    f.write(format_yaml_like(versions))