Merge pull request #35 from umccr/update/add-plotting-script-to-illum…

…ina-interop

Added plotting script to illumina interop container

repositories/illumina-interop/1.3.1/Dockerfile

@@ -5,14 +5,25 @@ LABEL author="Alexis Lucattini" \
       maintainer="[email protected]"
 
 ARG ILLUMINA_INTEROP_VERSION="1.3.1"
+ARG PANDAS_VERSION="2.2.2"
+ARG MATPLOTLIB_VERSION="3.9.0"
+ARG SEABORN_VERSION="0.13.2"
+
+COPY interop_imaging_plot.py /opt/conda/bin/interop_imaging_plot
 
 RUN conda config --prepend channels conda-forge && \
-    conda install -c bioconda illumina-interop=="${ILLUMINA_INTEROP_VERSION}" && \
+    conda install -y -c bioconda illumina-interop=="${ILLUMINA_INTEROP_VERSION}" && \
+    pip install --upgrade pip && \
+    pip install \
+        pandas=="${PANDAS_VERSION}" \
+        matplotlib=="${MATPLOTLIB_VERSION}" \
+        seaborn=="${SEABORN_VERSION}" && \
     find /opt/conda/ -follow -type f -name '*.a' -delete && \
     find /opt/conda/ -follow -type f -name '*.pyc' -delete && \
     /opt/conda/bin/conda clean --yes \
         --all \
-        --force-pkgs-dirs
+        --force-pkgs-dirs && \
+    chmod +x  /opt/conda/bin/interop_imaging_plot
 
 # Reset entrypoint to null for cwl
 ENTRYPOINT []

repositories/illumina-interop/1.3.1/interop_imaging_plot.py

@@ -0,0 +1,146 @@
+#!/usr/bin/env python3
+from functools import reduce
+
+# Imports
+import matplotlib.pyplot as plt
+import seaborn as sns
+import pandas as pd
+import sys
+import re
+
+from pathlib import Path
+from typing import Union, List
+
+
+"""
+Usage:
+
+interop_imaging_plot <input_csv> <output_png> <run_name>
+
+"""
+
+
+def header_regex_match(header_name) -> List[str]:
+    """
+    If the header name is in the format Name<item1;item2;item3>
+    :param header_name: 
+    :return: 
+    """
+    # Check if header name is in the format Name<item1;item2;item3>
+    header_regex = re.compile(r'(.*)<(.*)>')
+    header_match = header_regex.match(header_name)
+
+    if not header_match:
+        return [str(header_name)]
+
+    # Return the header name and the items
+    return list(
+        map(
+            lambda group_2_match_iter: f"{header_match.group(1)}_{group_2_match_iter}",
+            header_match.group(2).split(";")
+        )
+    )
+
+
+def read_csv(input_csv: Path) -> pd.DataFrame:
+    """
+    Read the input csv, and clean up the headers
+    :param input_csv:
+    :return:
+    """
+    # Read in header
+    # Headers % Base<A;C;G;T> should be
+    # Base_A, Base_C, Base_G, Base_T
+    # What an insane way to encode a header!!
+    imaging_df_headers_list = pd.read_csv(
+        input_csv,
+        comment='#',
+        header=0
+    ).columns.tolist()
+
+    imaging_df_headers_list = list(
+        reduce(
+            lambda x, y: x + y,
+            map(
+                lambda column_name_iter: list(header_regex_match(column_name_iter)),
+                imaging_df_headers_list
+            )
+        )
+    )
+
+    # Read in data
+    imaging_df = pd.read_csv(
+        input_csv,
+        # Skip comments
+        comment='#',
+        # Assign header but we overwrite it with names
+        header=0,
+        # Set our own header
+        names=imaging_df_headers_list
+    ).drop_duplicates(
+        subset=['Lane', '% Occupied', '% Pass Filter']
+    ).assign(
+        Lane=lambda row: row['Lane'].astype('category')
+    )
+
+    return imaging_df
+
+
+def plot_data(imaging_df: pd.DataFrame, output_png: Path, run_id: str) -> None:
+    """
+    Use the seaborn scatterplot library to plot the data
+    :param imaging_df:
+    :param output_png:
+    :param run_id:
+    :return:
+    """
+    # Write data
+    fig, ax = plt.subplots()
+
+    # Set grid style
+    sns.set_style('whitegrid')
+
+    # SNS Dot plot
+    sns.scatterplot(
+        x='% Occupied',
+        y='% Pass Filter',
+        data=imaging_df,
+        hue='Lane',
+        ax=ax,
+        alpha=0.6
+    )
+
+    # Set title
+    ax.set_title(f"Pct. Pass Filter vs. Pct. Occupied for run '{run_id}'")
+
+    # Set x-axis label
+    ax.set_xlabel('% Occupied')
+
+    # Set x limits
+    ax.set_xlim(left=0, right=100)
+
+    # Set y limits
+    ax.set_ylim(bottom=0, top=100)
+
+    # Set legend
+    ax.legend(title='Lane')
+
+    # Save plot
+    fig.savefig(sys.argv[2])
+
+
+def main():
+    # Set io
+    input_csv = sys.argv[1]
+    output_png = sys.argv[2]
+    run_id = sys.argv[3]
+
+    # Read in data
+    imaging_df = read_csv(Path(input_csv))
+
+    # Plot data
+    plot_data(imaging_df, Path(output_png), run_id)
+
+
+if __name__ == "__main__":
+    main()