Add patched MultiQC to the repo

.github/workflows/publish_multiqc.yml

@@ -0,0 +1,18 @@
+name: MultiQC to GitHub Container Registry
+
+on:
+  push:
+    branches: [main]
+    paths:
+      - "containers/multiqc/*"
+  workflow_dispatch:
+
+jobs:
+  call-workflow-passing-data:
+    uses: ./.github/workflows/reusable_publish_pypkg.yml
+    with:
+      package_name: multiqc
+      image: ${{ github.repository }}/multiqc
+      directory: ./containers/multiqc/
+    secrets:
+      GH_TOKEN: ${{ secrets.GH_TOKEN }}

.github/workflows/reusable_publish_pypkg.yml

@@ -0,0 +1,62 @@
+name: Reusable Workflow for Publish python packages to GitHub Container Registry
+
+on:
+  workflow_call:
+    inputs:
+      package_name:
+        required: true
+        type: string
+      registry:
+        required: false
+        type: string
+        default: ghcr.io
+      image:
+        required: true
+        type: string
+      directory:
+        required: true
+        type: string
+    secrets:
+      GH_TOKEN:
+        required: true
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Check out repository
+        uses: actions/checkout@v4
+
+      - name: Log into Container Registry
+        uses: docker/login-action@v3
+        with:
+          registry: ${{ inputs.registry }}
+          username: ${{ github.actor }}
+          password: ${{ secrets.GH_TOKEN }}
+
+      - name: Set up Docker Buildx
+        id: buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Get version
+        id: getversion
+        run: |
+          PACKAGE=${{ inputs.package_name }}
+          DIRECTORY=${{ inputs.directory }}
+          VERSION=$(cat ./${DIRECTORY}/requirements.txt | grep -Po "(?<=${PACKAGE}==)[0-9.]+")
+          echo "version=$VERSION" >> $GITHUB_OUTPUT
+
+      - name: Build and push
+        id: docker_build
+        uses: docker/build-push-action@v6
+        with:
+          context: ${{ inputs.directory }}
+          file: ${{ inputs.directory }}/Containerfile
+          push: true
+          tags: |
+            ${{ inputs.registry }}/${{ inputs.image }}:${{ steps.getversion.outputs.version }}
+            ${{ inputs.registry }}/${{ inputs.image }}:latest
+
+      - name: Image digest
+        run: echo ${{ steps.docker_build.outputs.digest }}

containers/multiqc/Containerfile

@@ -0,0 +1,24 @@
+FROM python:3.11-slim
+
+ARG DEBIAN_FRONTEND="noninteractive"
+
+LABEL author="Diego Alvarez ([email protected])"
+LABEL description="MultiQC (https://github.com/MultiQC/MultiQC) with nanoq patched module"
+LABEL org.opencontainers.image.source="https://github.com/dialvarezs/containers"
+
+RUN apt-get update && apt-get upgrade -y && \
+    apt-get install -y procps patch && \
+    rm -fr /var/lib/apt/lists/*
+
+ENV VIRTUAL_ENV=/venv
+ENV PATH="${VIRTUAL_ENV}/bin:$PATH"
+
+COPY requirements.txt /tmp/requirements.txt
+RUN --mount=from=ghcr.io/astral-sh/uv:latest,source=/uv,target=/bin/uv \
+    uv venv ${VIRTUAL_ENV} && \
+    uv pip install --no-cache -r /tmp/requirements.txt
+
+# patch nanoq module ot use bar chart instead of line chart
+COPY nanoq_bargraph.patch /tmp/nanoq_bargraph.patch
+RUN cd /venv/lib/python3.11/site-packages/multiqc/ && \
+    patch -p1 < /tmp/nanoq_bargraph.patch 

containers/multiqc/nanoq_bargraph.patch

@@ -0,0 +1,132 @@
+diff --git a/modules/nanoq/nanoq.py b/modules/nanoq/nanoq.py
+index 398b2a24..bbc0d711 100644
+--- a/modules/nanoq/nanoq.py
++++ b/modules/nanoq/nanoq.py
+@@ -5,8 +5,9 @@ from copy import deepcopy
+ from typing import Dict, List, Any
+
+ from multiqc.base_module import BaseMultiqcModule, ModuleNoSamplesFound
+-from multiqc.plots import table, linegraph
++from multiqc.plots import table, bargraph, linegraph
+ from multiqc.plots.table_object import TableConfig
++from multiqc.utils import mqc_colour
+
+ log = logging.getLogger(__name__)
+
+@@ -131,7 +132,7 @@ class MultiqcModule(BaseMultiqcModule):
+         )
+
+         general_stats_headers = deepcopy(headers)
+-        for k, h in general_stats_headers.items():
++        for h in general_stats_headers.values():
+             h["hidden"] = True
+         general_stats_headers["Number of reads"]["hidden"] = False
+         general_stats_headers["N50 read length"]["hidden"] = False
+@@ -141,16 +142,39 @@ class MultiqcModule(BaseMultiqcModule):
+
+     def reads_by_quality_plot(self, data_by_sample: Dict[str, Dict[str, float]]) -> None:
+         # Get data for plot
+-        lineplot_data: Dict[str, Dict[int, float]] = defaultdict(dict)
++        barplot_data: Dict[str, Dict[str, float]] = defaultdict(dict)
++        min_quality = 10
++
+         for name, d in data_by_sample.items():
+-            reads_by_q = {k: v for k, v in d.items() if k.startswith("Reads > Q")}
+-            if len(reads_by_q) == 0:
++            reads_by_q = {
++                int(k.split("> Q")[1]): v for k, v in d.items() if k.startswith("Reads > Q")
++            }
++            if not reads_by_q:
++                continue
++
++            thresholds = sorted(th for th in reads_by_q if th >= min_quality)
++            if not thresholds:
+                 continue
+
+-            total_reads = d["Number of reads"]
+-            for k, v in reads_by_q.items():
+-                threshold = int(k.split("> Q")[1])
+-                lineplot_data[name][threshold] = v / total_reads * 100
++            keys = [f"<Q{min_quality}"]
++
++            for th, thn in zip(thresholds[:-1], thresholds[1:]):
++                key = f"Q{th}-{thn}"
++                keys.append(key)
++                barplot_data[name][key] = reads_by_q[th] - reads_by_q[thn]
++
++            last_key = f">Q{thresholds[-1]}"
++            keys.append(last_key)
++            barplot_data[name][f"<Q{min_quality}"] = (
++                d["Number of reads"] - reads_by_q[thresholds[0]]
++            )
++            barplot_data[name][last_key] = reads_by_q[thresholds[-1]]
++
++        colours = mqc_colour.mqc_colour_scale("RdYlGn-rev", 0, len(keys))
++        cats = {
++            k: {"name": f"Reads {k}", "color": colours.get_colour(idx, lighten=1)}
++            for idx, k in enumerate(keys[::-1])
++        }
+
+         # Plot
+         self.add_section(
+@@ -163,9 +187,10 @@ class MultiqcModule(BaseMultiqcModule):
+             The phred score represents the liklelyhood that a given read contains errors.
+             High quality reads have a high score.
+             """,
+-            plot=linegraph.plot(
+-                lineplot_data,
+-                linegraph.LinePlotConfig(
++            plot=bargraph.plot(
++                barplot_data,
++                cats,
++                pconfig=dict(
+                     id="nanoq_plot_quality_plot",
+                     title="Nanoq: read qualities",
+                 ),
+@@ -173,6 +198,7 @@ class MultiqcModule(BaseMultiqcModule):
+         )
+
+     def reads_by_length_plot(self, data_by_sample: Dict[str, Dict[str, float]]) -> None:
++        # sourcery skip: simplify-len-comparison
+         # Get data for plot
+         linegraph_data: Dict[str, Dict[int, float]] = defaultdict(dict)
+         for name, d in data_by_sample.items():
+@@ -235,12 +261,18 @@ def parse_nanoq_log(f) -> Dict[str, float]:
+
+     # Helper function to parse thresholds part
+     def parse_thresholds(lines: List[str], threshold_type: str) -> Dict[str, List[Any]]:
+-        _thresholds: Dict[str, List[Any]] = {"Threshold": [], "Number of Reads": [], "Percentage": []}
++        _thresholds: Dict[str, List[Any]] = {
++            "Threshold": [],
++            "Number of Reads": [],
++            "Percentage": [],
++        }
+         for _line in lines:
+             match = re.match(r">\s*(\d+)\s+(\d+)\s+(\d+\.\d+)%", _line)
+             if match:
+                 _threshold, _num_reads, _percentage = match.groups()
+-                _thresholds["Threshold"].append(_threshold + (threshold_type if threshold_type == "bp" else ""))
++                _thresholds["Threshold"].append(
++                    _threshold + (threshold_type if threshold_type == "bp" else "")
++                )
+                 _thresholds["Number of Reads"].append(int(_num_reads))
+                 _thresholds["Percentage"].append(float(_percentage))
+         return _thresholds
+@@ -248,10 +280,15 @@ def parse_nanoq_log(f) -> Dict[str, float]:
+     read_length_thresholds = parse_thresholds(length_threshold_lines, "bp")
+     read_quality_thresholds = parse_thresholds(quality_threshold_lines, "")
+
+-    for threshold, num_reads in zip(read_length_thresholds["Threshold"], read_length_thresholds["Number of Reads"]):
++    for threshold, num_reads in zip(
++        read_length_thresholds["Threshold"], read_length_thresholds["Number of Reads"]
++    ):
+         stats[f"Reads > {threshold}"] = num_reads
+
+-    for threshold, num_reads in zip(read_quality_thresholds["Threshold"], read_quality_thresholds["Number of Reads"]):
++    for threshold, num_reads in zip(
++        read_quality_thresholds["Threshold"], read_quality_thresholds["Number of Reads"]
++    ):
+         stats[f"Reads > Q{threshold}"] = num_reads
+
+     return stats
++

containers/multiqc/requirements.txt

@@ -0,0 +1 @@
+multiqc==1.24.1