Merge pull request #12 from Russel88/dev

v.0.1.20 added feature for detecting strain-mixtures using allele frequeicies

maginator/main.py

@@ -78,6 +78,7 @@ def cli():
     app.add_argument('--min_nonN', help='Minimum fraction of non-N characters of a sample to be included in a phylogeny [%(default)s]', default=0.5, type=float)
     app.add_argument('--min_marker_genes', help='Minimum marker genes to be detected for inclusion of a sample in a phylogeny [%(default)s]', default=2, type=int)
     app.add_argument('--min_signature_genes', help='Minimum signature genes to be detected for inclusion of a sample in a phylogeny [%(default)s]', default=50, type=int)
+    app.add_argument('--af_cutoff', help='Cutoff for average median allele frequency of SG alignment for determining mixed/pure population of a MAG in a sample [%(default)s]', default=0, type=float)
     app.add_argument('--phylo', help='Software for phylogeny inference. Either fast (fasttree) or slow and more accurate (iqtree) [%(default)s]', default='fasttree', type=str, choices=['fasttree', 'iqtree'])
     app.add_argument('--tax_scope_threshold', help='Threshold for assigning the taxonomic scope of a gene cluster [%(default)s]', default=0.9, type=float)
     app.add_argument('--synteny_adj_cutoff', help='Minimum number of times gene clusters should be adjacent to be included in synteny graph [%(default)s]', default=1, type=int)

maginator/workflow/pileup_parse.Snakefile

@@ -42,3 +42,22 @@ rule parse:
     script:
         "scripts/mpileup.py"
 
+rule allele_frequencies:
+    input:
+        stat=os.path.join(WD, 'phylo', 'stats.tab'),
+        gene=os.path.join(WD, 'phylo', 'stats_genes.tab')
+    output:
+        af_matrix=os.path.join(WD,'tabs','allele_frequencies.tab')
+    params:
+        af_cutoff = param_dict['af_cutoff'],
+        min_signature_genes = param_dict['min_signature_genes'],
+        min_nonN = param_dict['min_nonN'],
+        min_marker_genes = param_dict['min_marker_genes']
+    conda:
+        "envs/signature_genes.yaml"
+    resources:
+        cores=2,
+        mem_gb=190,
+        runtime='172800'
+    script:
+        "scripts/allele_frequency_mat.R"

maginator/workflow/scripts/allele_frequency_mat.R

@@ -0,0 +1,59 @@
+library(dplyr)
+gene_stats_file <- snakemake@input[["gene"]]
+gene_stats <- read.table(file=gene_stats_file, header=TRUE)
+
+stats_file <- snakemake@input[["stat"]]
+stats <- read.table(file=stats_file, header=TRUE)
+
+af_cutoff <- as.numeric(snakemake@params[["af_cutoff"]])
+min_signature_genes <- as.integer(snakemake@params[["min_signature_genes"]])
+min_nonN <- as.numeric(snakemake@params[["min_nonN"]])
+min_marker_genes <- as.integer(snakemake@params[["min_marker_genes"]])
+
+# Identifying the samples in each tree fulfill the default tree-criteria
+#(<50 detected SG, min 2 marker genes, min 50% non-n coverage)
+filtered_data <- stats %>%
+  filter(Marker_N >= min_marker_genes, Signature_N>=min_signature_genes, NonN > min_nonN)
+
+# Count the number of different samples for each cluster
+sample_counts <- filtered_data %>%
+  group_by(Cluster) %>%
+  summarize(Count = n_distinct(Sample))
+
+# Get unique clusters
+unique_clusters <- unique(gene_stats$Cluster)
+unique_samples <- unique(gene_stats$Sample)
+
+# Create an empty matrix
+empty_matrix <- matrix(NA, nrow = length(unique_samples), ncol = length(unique_clusters),
+                       dimnames = list(unique_samples, unique_clusters))
+
+# Loop through each cluster
+for (cluster_id in unique_clusters) {
+  if ((dim(gene_stats[gene_stats$Cluster == cluster_id,]))[1]>af_cutoff){
+    # Subset data for the current cluster
+    cluster_data <- gene_stats[gene_stats$Cluster == cluster_id, ]
+    samples <- (filtered_data[filtered_data$Cluster==cluster_id, "Sample"])
+    absent_samples <- unique_samples[!unique_samples %in% samples]
+
+    # Calculate the median Signature_AF per sample for the current cluster
+    median_signature_af_per_sample <- cluster_data %>%
+      group_by(Sample) %>%
+      summarise(median_signature_af = median(Signature_AF, na.rm = TRUE))
+
+    # Convert median_signature_af_per_sample to a data frame
+    median_signature_af_per_sample_df <- as.data.frame(median_signature_af_per_sample)
+    # Replace values based on conditions
+    median_signature_af_per_sample_df$median_signature_af[median_signature_af_per_sample_df$median_signature_af > af_cutoff] <- 2
+    median_signature_af_per_sample_df$median_signature_af[median_signature_af_per_sample_df$median_signature_af == af_cutoff] <- 1
+
+    # Filling in the numbers in the matrix
+    empty_matrix[,cluster_id] = median_signature_af_per_sample_df$median_signature_af
+    empty_matrix[absent_samples,cluster_id] = 0 # if the sample is not in the tree - set NA as AF
+  }
+}
+
+AF_matrix <- empty_matrix
+
+# Save the data frame to a CSV file
+write.csv(AF_matrix, file = snakemake@output[["af_matrix"]], row.names = TRUE)

setup.py

@@ -5,7 +5,7 @@
 
 setuptools.setup(
     name="maginator", 
-    version="0.1.19",
+    version="0.1.20",
     author="Jakob Russel & Trine Zachariasen",
     author_email="[email protected],[email protected]",
     description="MAGinator: Abundance, strain, and functional profiling of MAGs",