Add bcftools stats for a vcfFile raw files

tools/IBS_gds/IBS_gds.script.R

@@ -25,13 +25,13 @@ ibs.hc <- snpgdsHCluster(ibs)
     {
         pop_map= read.table(popmap_file, header=F, sep='\t') ; 
         colnames(pop_map)[c(1,2)]=c('sample.id', 'popcode') ; 
-        pops = factor(pop_map$popcode)[match(ibsca$sample.id, pop_map$sample.id)]
+        pops = factor(pop_map$popcode)[match(ibs$sample.id, pop_map$sample.id)]
 
         # individulas in the same population are clustered together
         pop.idx <- order(pops)
 
         # Determine groups of individuals by population information
-        rv2 <- snpgdsCutTree(ibs.hc, samp.group=pops)
+        rv <- snpgdsCutTree(ibs.hc, samp.group=pops)
 
     } else {
         pop.idx = seq(1, length(ibs$sample.id))
@@ -57,8 +57,7 @@ png(file=ibs_mdscale_png, width=1024, height=1024);
     {
         plot(x, y, col=pops, xlab = "", ylab = "",  main = "Multidimensional Scaling Analysis (on IBS distances)")
         legend("topleft", legend=levels(pops), pch="o", text.col=1:nlevels(pops))
-    }
-    else{
+    }    else{
         plot(x, y, xlab = "", ylab = "",  main = "Multidimensional Scaling Analysis (on IBS distances)")
     }
 dev.off()

tools/IBS_gds/IBS_gds.yaml

@@ -19,7 +19,7 @@
                  { name: popmap_file, type: "popmap", file: "", description: "Path to tsv file with samples group"}
                 ],
         outputs: [
-          { name: ibs, type: "txt", file: "*_k0.txt", description: "a matrix of IBS proportion" },
+          { name: ibs, type: "txt", file: "*_ibs.txt", description: "a matrix of IBS proportion" },
           { name: ibs_dendro_png, type: "png", file: "ibs_dendro_plot_mqc.png", description: "HClust and determine groups of individuals by population information or automatically" },
           { name: ibs_heatmap_png, type: "png", file: "ibs_heatmap_mqc.png", description: "A heatmap of IBS pairwise identities" },
           { name: ibs_mdscale_png, type: "png", file: "ibs_mdscale_mqc.png", description: "Multidimensional scaling analysis on the   matrix of genome-wide IBS pairwise distances" },

tools/bcftools_vcfFile_stats/bcftools_vcfFile_stats.rule.snakefile

+rule <step_name>__bcftools_vcfFile_stats:
+    input:
+        **<step_name>__bcftools_vcfFile_stats_inputs(),
+    output: 
+        stats = config["results_dir"] + "/" + config["<step_name>__bcftools_vcfFile_stats_output_dir"] + "/Allsamples_vcf_stats.txt",
+        stats_image = config["results_dir"] + "/" + config["<step_name>__bcftools_vcfFile_stats_output_dir"] + "/Allsamples_stats_mqc.png",
+    params: 
+        output_dir = config["results_dir"] + "/" + config["<step_name>__bcftools_vcfFile_stats_output_dir"]+ "/Allsamples",
+        command = config["<step_name>__bcftools_vcfFile_stats_command"],
+    log: 
+        config["results_dir"] + "/logs/" + config["<step_name>__bcftools_vcfFile_stats_output_dir"] + "/Allsamples_bcftools_vcfFile_stats_log.txt"
+    shell: 
+        "bcftools merge {input.vcf} |  bgzip > {params.output_dir}.vcf.gz ; "
+        "{params.command} "
+        "-s - "
+        "{params.output_dir}.vcf.gz "
+        "> {output.stats} "
+        "2> {log}; "
+        "plot-vcfstats "
+        "-p {params.output_dir} "
+        "-P " # no pdf
+        "{output.stats}; "
+        # Merge images
+        "convert -size 1200x1000 \( {params.output_dir}/depth.0.png  {params.output_dir}/substitutions.0.png {params.output_dir}/hwe.0.png +smush +400 \) "
+        "\( {params.output_dir}/indels_by_sample.0.png {params.output_dir}/snps_by_sample.0.png +smush +150 \) "
+        "\( {params.output_dir}/dp_by_sample.0.png {params.output_dir}/singletons_by_sample.0.png +smush +150 \) "
+        "\( {params.output_dir}/hets_by_sample.0.png {params.output_dir}/tstv_by_sample.0.png +smush +150 \) -append {output.stats_image}"
\ No newline at end of file

tools/bcftools_vcfFile_stats/bcftools_vcfFile_stats.yaml

+{
+  id: "bcftools_vcfFile_stats",
+  name: "BCFtools VCF Stats",
+  description: "BCFtools is a set of utilities that manipulate variant calls in the Variant Call Format (VCF) and its binary counterpart BCF.",
+  version: "1.9",
+  website: "https://samtools.github.io/bcftools/",
+  git: "https://github.com/samtools/bcftools",
+  documentation: "https://samtools.github.io/bcftools/howtos/index.html",
+  article: "10.1093/bioinformatics/btr509",
+  multiqc: "custom",
+  commands:
+    [
+      {
+        name: "bcftools_vcfFile_stats",
+        cname: "BCFtools VCFFile Stats",
+        command: "bcftools stats",
+        category: vcf_postprocess,
+        output_dir: "bcftools_vcfFile_stats",
+        inputs:
+          [
+            {
+              name: "vcf",
+              type: "vcfFile",
+              file: "*_variants.vcf.gz",
+              description: "Variant file (vcf)",
+            },
+          ],
+        outputs:
+          [
+            {
+              name: "stats",
+              type: "txt",
+              file: "Allsamples_vcf_stats.txt",
+              description: "VCF file stats",
+            },
+            {
+              name: "stats_image",
+              type: "png",
+              file: "Allsamples_stats_mqc.png",
+              description: "VCF file stats graphs",
+            }
+          ],
+        options:
+          [
+          ],
+      },
+    ],
+  install:
+    {
+      bcftools:
+      [
+        "cd /opt/biotools",
+        "wget https://github.com/samtools/bcftools/releases/download/1.9/bcftools-1.9.tar.bz2",
+        "tar -xvjf bcftools-1.9.tar.bz2",
+        "cd bcftools-1.9",
+        "./configure --prefix=/opt/biotools",
+        "make -j 10",
+        "make install",
+        "mv bcftools /opt/biotools/bin/",
+        "cd .. && rm -r bcftools-1.9.tar.bz2 bcftools-1.9",
+        "sed -i 's/python/python3/g' /opt/biotools/bin/plot-vcfstats",
+        "sed -i 's|11/2.54|20/2.54|' /opt/biotools/bin/plot-vcfstats",
+        "sed -i 's|10/2.54|14/2.54|' /opt/biotools/bin/plot-vcfstats",
+        "sed -i 's|window_len/2|window_len//2|g' /opt/biotools/bin/plot-vcfstats"
+      ],
+      tabix:
+      [
+        "apt -y update && apt install -y tabix"
+      ],
+      matplotlib:
+      [
+        "pip3 install matplotlib"
+      ],
+      imagemagick:
+      [
+        "apt -y update && apt install -y imagemagick"
+      ],
+      bcftools_dependencies: [
+        "apt -y update && apt install -y python-backports.functools-lru-cache"
+      ]
+    },
+  citations:  {
+    bcftools: [
+      "Heng Li, A statistical framework for SNP calling, mutation discovery, association mapping and population genetical parameter estimation from sequencing data, Bioinformatics, Volume 27, Issue 21, 1 November 2011, Pages 2987-2993, https://doi.org/10.1093/bioinformatics/btr509"
+    ],
+    tabix: [
+      "Heng Li, Tabix: fast retrieval of sequence features from generic TAB-delimited files, Bioinformatics, Volume 27, Issue 5, 1 March 2011, Pages 718-719, https://doi.org/10.1093/bioinformatics/btq671"
+    ],
+    matplotlib: [
+      "J. D. Hunter, 'Matplotlib: A 2D Graphics Environment,' in Computing in Science & Engineering, vol. 9, no. 3, pp. 90-95, May-June 2007, doi: 10.1109/MCSE.2007.55."
+    ],
+  }
+}