Relax fallback threshold

bin/plot_metagenomes.py

@@ -64,7 +64,7 @@ def __init__(
         )
         self.adata.var = self.adata.var.assign(**{
             kw: self.adata.var[kw].fillna(val)
-            for kw, val in [("estimate", 0), ("std_error", 0), ("p_value", 1)]
+            for kw, val in [("estimate", 0), ("std_error", 0), ("p_value", 0.99)]
         })
         self.adata.var["qvalue"] = multipletests(self.adata.var["p_value"].fillna(0), 0.1, "fdr_bh")[1]
         self.adata.var["neg_log10_pvalue"] = -np.log10(self.adata.var["p_value"])
@@ -244,7 +244,7 @@ def plot(self, meta_cname, output_folder):
 
         # Bars showing the -log10(qvalue) for each gene bin
         # (unless there are no useful q-values, in which case fall back to p-values)
-        if self.adata.var["neg_log10_qvalue"].max() == 0:
+        if self.adata.var["neg_log10_qvalue"].max() < 0.0001:
             kw = "neg_log10_pvalue"
             label = "p-value (-log10)"
         else: