add eps from clustering to config

README.md

@@ -25,6 +25,9 @@ alternative_haplotype_vcf: # str
 # path to liftover directory (explained below)
 liftover_root: # str, defaults to 'resources/references/liftover'
 
+clustering:
+  eps: 0.6 # float, the higher the bigger the clusters can be
+
 # Settings for the pairwise alignment
 pw_alignment:
   match: # positive integer

config/config.yaml

@@ -11,6 +11,9 @@ pw_alignment:
   gap: -1
   mismatch: -1
 
+clustering:
+  eps: 0.6 # float, the higher the bigger the clusters can be
+
 samples: ['17_08']
 
 # display names and file names of the used references

workflow/Snakefile

@@ -108,9 +108,11 @@ rule cluster:
         sequences = "results/{sample}/{region}/sequences.fa"
     output:
         clustering = "results/{sample}/{region}/clustering.txt"
+    params:
+        eps = config['clustering']['eps']
     shell:
         """
-            python workflow/scripts/read_cluster.py cluster {input.sequences} {input.dists} -o {output.clustering}
+            python workflow/scripts/read_cluster.py cluster {input.sequences} {input.dists} --eps {params.eps} -o {output.clustering}
         """
 
 rule pca_plot:

workflow/scripts/read_cluster.py

@@ -213,8 +213,8 @@ def parse_args():
     cluster_parser = subparsers.add_parser('cluster', help="Cluster reads by distance.")
     cluster_parser.add_argument("sequences", type=Path)
     cluster_parser.add_argument("distances", type=Path)
-    cluster_parser.add_argument("--eps", type=float, default=0.15)
     cluster_parser.add_argument("--min_samples", type=int, default=5)
+    cluster_parser.add_argument("--eps", type=float, default=0.3)
     cluster_parser.add_argument("--output", "-o", type=Path, default=None)
 
     plot_parser = subparsers.add_parser('plot', help="Plot the clustering in a PCA plot (2D).")