Better documentation and debug error in plotting allele stats for 0 a…

…lleles
pinellolab · May 6, 2024 · f20315d · f20315d
1 parent 9024cae
commit f20315d
Show file tree

Hide file tree

Showing 4 changed files with 5 additions and 3 deletions.
diff --git a/bean/framework/ReporterScreen.py b/bean/framework/ReporterScreen.py
@@ -621,6 +621,7 @@ def get_normalized_allele_counts(self, allele_count_df=None, norm_thres=10):
         norm_counts = self._get_allele_norm(
             allele_count_df=allele_count_df, thres=norm_thres
         )
+        alleles_norm.loc[:, count_columns] = alleles_norm.loc[:, count_columns].astype(float)
         alleles_norm.loc[:, count_columns] = (
             alleles_norm.loc[:, count_columns] / norm_counts
         )

diff --git a/bean/model/parser.py b/bean/model/parser.py
@@ -83,7 +83,7 @@ def parse_args(parser=None):
         "--control-condition",
         default="bulk",
         type=str,
-        help="Value in `bdata.samples[condition_col]` that indicates control experimental condition whose editing patterns will be used.",
+        help="Value in `bdata.samples[condition_col]` that indicates control experimental condition whose editing patterns will be used. Select this as the condition with the least selection- For the sorting screen, use presort (bulk). For the survival screens, use the closest one with T=0.",
     )
     parser.add_argument(
         "--exclude-control-condition-for-inference",

diff --git a/bean/model/run.py b/bean/model/run.py
@@ -99,7 +99,7 @@ def check_args(args, bdata):
         )
     if args.control_condition not in bdata.samples[args.condition_col].tolist():
         raise ValueError(
-            f"No sample has control label `{args.control_condition}` (set by `--control-condition`)  in ReporterScreen.samples[{args.condition_col}]: {bdata.samples[args.condition_col]}. Check your input."
+            f"No sample has control label `{args.control_condition}` (set by `--control-condition`)  in ReporterScreen.samples[{args.condition_col}]: {bdata.samples[args.condition_col]}. Check your input.  For the selection of this argument, see more in `--condition-col` under `bean run --help`."
         )
     if args.replicate_col not in bdata.samples.columns:
         raise ValueError(

diff --git a/bean/plotting/allele_stats.py b/bean/plotting/allele_stats.py
@@ -39,7 +39,8 @@ def plot_n_guides_per_edit(
     n_guides = edits_df.groupby("edits")["guide"].count()
     if ax is None:
         fig, ax = plt.subplots()
-    ax.hist(n_guides, bins=np.arange(min(n_guides) - 0.5, max(n_guides) + 0.5))
+    if len(n_guides) > 0:
+        ax.hist(n_guides, bins=np.arange(min(n_guides) - 0.5, max(n_guides) + 0.5))
     ax.set_xlabel("# guides per variant")
     ax.set_title(f"n_variants={len(edits_df)}")
     ax.set_ylabel("# edits")