Watchdog update, jbrowse mode

workflow/rules/aggregate_fct.smk

@@ -254,6 +254,7 @@ def selected_input_bam(wildcards):
     )
 
 
+
 def selected_input_bai(wildcards):
     """
     Function based on checkpoint filter_bad_cells_from_mosaic_count

workflow/rules/common.smk

@@ -441,6 +441,11 @@ if config["scNOVA"] is True:
         labels_path = "{folder}/{sample}/cell_selection/labels.tsv".format(
             folder=config["data_location"], sample=sample
         )
+
+        assert os.path.isfile(
+            labels_path
+        ), "Ashleys labels were not computed yet, use first ashleys mode to perform cell selection"
+
         # print(labels_path)
         if os.path.exists(labels_path):
             # Read df
@@ -843,6 +848,20 @@ def get_all_plots(wildcards):
                     sample=wildcards.sample,
                 ),
             )
+            l_outputs.extend(
+                expand(
+                    "{folder}/{sample}/plots/UCSC/{sample}.bedUCSC.gz",
+                    folder=config["data_location"],
+                    sample=wildcards.sample,
+                ),
+            )
+            l_outputs.extend(
+                expand(
+                    "{folder}/{sample}/plots/JBROWSE/{sample}.ok",
+                    folder=config["data_location"],
+                    sample=wildcards.sample,
+                ),
+            )
 
         # Stats section
 

workflow/rules/plots.smk

@@ -334,6 +334,27 @@ rule scTRIP_multiplot_aggr:
         mem_mb=get_mem_mb,
 
 
+rule jbrowse_genome_browser_file:
+    input:
+        counts="{folder}/{sample}/counts/{sample}.txt.gz",
+        stringent_calls=(
+            "{folder}/{sample}/mosaiclassifier/sv_calls/stringent_filterTRUE.tsv"
+        ),
+    output:
+        "{folder}/{sample}/plots/JBROWSE/{sample}.ok",
+    log:
+        "{folder}/log/JBROWSE/{sample}.log",
+    conda:
+        "/g/korbel2/weber/miniconda3/envs/genome_browsing"
+        # "../envs/genome_browsing.yaml"
+    container:
+        None
+    resources:
+        mem_mb=get_mem_mb,
+    shell:
+        "python workflow/scripts/genome_browsing/generate_jbrowse_tracks.py {input.counts} {input.stringent_calls} {output} > {log}"
+
+
 rule ucsc_genome_browser_file:
     input:
         counts="{folder}/{sample}/counts/{sample}.txt.gz",
@@ -346,7 +367,7 @@ rule ucsc_genome_browser_file:
     output:
         "{folder}/{sample}/plots/UCSC/{sample}.bedUCSC.gz",
     log:
-        "{folder}/log/ucsc_genome_browser_file/{sample}.bedUCSC.gz",
+        "{folder}/log/ucsc_genome_browser_file/{sample}.bedUCSC.log",
     conda:
         "../envs/mc_base.yaml"
     resources:

workflow/scripts/genome_browsing/generate_jbrowse_tracks.py

@@ -0,0 +1,113 @@
+import pyBigWig
+import pandas as pd
+import sys, os
+import subprocess
+
+# Assuming sv_cell_df already contains the 'sv_call_name' column
+
+
+# Function to map sv_call_name to color
+def map_color(sv_call_name):
+    colors = {
+        "none": "#F8F8F8",
+        "del_h1": "#77AADD",
+        "del_h2": "#4477AA",
+        "del_hom": "#114477",
+        "dup_h1": "#CC99BB",
+        "dup_h2": "#AA4488",
+        "dup_hom": "#771155",
+        "inv_h1": "#DDDD77",
+        "inv_h2": "#AAAA44",
+        "inv_hom": "#777711",
+        "idup_h1": "#DDAA77",
+        "idup_h2": "#AA7744",
+        "complex": "#774411",
+    }
+    return colors.get(sv_call_name, "#000000")  # Default to black if not found
+
+
+# Load read counts data
+counts_file_init = pd.read_csv(sys.argv[1], sep="\t", compression="gzip")
+chrom_size_df = counts_file_init.groupby("chrom")["end"].max().reset_index()
+
+# Load SV data
+sv_df = pd.read_csv(sys.argv[2], sep="\t")
+
+output_file = sys.argv[3]
+output_dir = "/".join(output_file.split("/")[:-1])
+
+os.makedirs(output_dir, exist_ok=True)
+
+
+# Process each cell
+for cell in sorted(counts_file_init.cell.unique()):
+    print(cell)
+    # Filter read counts for current cell
+    counts_file = counts_file_init[counts_file_init["cell"] == cell]
+    print(counts_file)
+    counts_file["w"] = counts_file["w"].astype(float)
+    counts_file["c"] = counts_file["c"].astype(float) * -1
+
+    # Create BigWig for Watson counts
+    bw = pyBigWig.open(f"{output_dir}/{cell}-W.bigWig", "w")
+    bw.addHeader(list(chrom_size_df.itertuples(index=False, name=None)))
+    bw.addEntries(
+        counts_file.chrom.values.tolist(),
+        counts_file.start.values.tolist(),
+        ends=counts_file.end.values.tolist(),
+        values=counts_file.w.values.tolist(),
+    )
+    bw.close()
+
+    # Create BigWig for Crick counts
+    bw = pyBigWig.open(f"{output_dir}/{cell}-C.bigWig", "w")
+    bw.addHeader(list(chrom_size_df.itertuples(index=False, name=None)))
+    bw.addEntries(
+        counts_file.chrom.values.tolist(),
+        counts_file.start.values.tolist(),
+        ends=counts_file.end.values.tolist(),
+        values=counts_file.c.values.tolist(),
+    )
+    bw.close()
+
+    # Process SV data for current cell
+    sv_cell_df = sv_df[sv_df["cell"] == cell]
+    # sv_cell_df = sv_cell_df[["chrom", "start", "end", "sv_call_name"]]
+
+    # Add a color column
+    sv_cell_df["color"] = sv_cell_df["sv_call_name"].apply(map_color)
+
+    # Select the relevant columns for the BED file
+    sv_cell_df = sv_cell_df[
+        [
+            "chrom",
+            "start",
+            "end",
+            "sv_call_name",
+            "color",
+            "sv_call_haplotype",
+            "llr_to_ref",
+            "af",
+        ]
+    ]
+
+    # Write to BED file
+    sv_filename = f"{output_dir}/{cell}-SV.bed"
+    sv_cell_df.to_csv(sv_filename, sep="\t", index=False, header=False)
+
+    # Compress the file using bgzip
+    compressed_filename = sv_filename + ".gz"
+    subprocess.run(["bgzip", "-c", sv_filename], stdout=open(compressed_filename, "wb"))
+
+    # Index the compressed file using tabix
+    subprocess.run(["tabix", "-p", "bed", compressed_filename])
+
+    # Create BigWig for SV - this step might need adjustment based on SV data format
+    # bw = pyBigWig.open(sv_filename, "w")
+    # bw.addHeader(list(chrom_size_df.itertuples(index=False, name=None)))
+    # Additional logic for adding SV data to BigWig may be required here
+    # bw.close()
+
+from pathlib import Path
+
+Path(output_file).touch()

workflow/scripts/strandphaser_scripts/prepare_strandphaser.py

@@ -2,7 +2,15 @@
     print("[General]", file=f)
     print("numCPU           = 1", file=f)
     print("chromosomes      = '" + snakemake.wildcards.chrom + "'", file=f)
-    print("pairedEndReads   = '" + [e.strip() for e in open(snakemake.input.single_paired_end_detect, "r").readlines()][0] + "'", file=f)
+    print(
+        "pairedEndReads   = '"
+        + [
+            e.strip()
+            for e in open(snakemake.input.single_paired_end_detect, "r").readlines()
+        ][0]
+        + "'",
+        file=f,
+    )
     print("min.mapq         = 10", file=f)
     print("", file=f)
     print("[StrandPhaseR]", file=f)
@@ -16,5 +24,13 @@
     print("compareSingleCells = FALSE", file=f)
     print("callBreaks       = FALSE", file=f)
     print("exportVCF        = '", snakemake.wildcards.sample, "'", sep="", file=f)
-    print("bsGenome         = '", snakemake.config["references_data"][snakemake.config["reference"]]["R_reference"], "'", sep="", file=f)
+    print(
+        "bsGenome         = '",
+        snakemake.config["references_data"][snakemake.config["reference"]][
+            "R_reference"
+        ],
+        "'",
+        sep="",
+        file=f,
+    )
     # print("bsGenome         = '", snakemake.config["R_reference"], "'", sep="", file=f)