diff --git a/README.md b/README.md
index 67bfe14..e055b5f 100644
--- a/README.md
+++ b/README.md
@@ -130,23 +130,24 @@ When the script finishes, it creates `graph.pkl` in the specified output
 directory. `graph.pkl` is a Python pickle file representing the constructed
 gene matches graph.
 
-The script also creates Python pickles for the pairwise BLAST results. The BLAST
-results can be found in the `od2` subdirectory of the output directory.
+The script also stores HDF5 files (formerly Python pickles) for the pairwise
+BLAST results. The BLAST results can be found in the `od2` subdirectory of the
+output directory.
 
 ### Phase 2: Calculating distances
 
 The `filtered_distance.py` Python script may be used to compute distances or
 similarities from a gene matches graph. Basic usage of the command requires
-only that we provide the pickles for the gene matches graph and the pairwise
-BLAST results.
+only that we provide the pickles for the gene matches graph and the HDF5 files
+for the pairwise BLAST results.
 
 ```bash
-python filtered_distance.py -g GRAPH -c COMPARISONS_DIR/*.pkl
+python filtered_distance.py -g GRAPH -c COMPARISONS_DIR/*.h5
 ```
 
 In the above command, GRAPH should be the path to the `graph.pkl` created in the
 first phase, and COMPARISONS_DIR should be the directory that contains the BLAST
-result pickles. (This will be the `od2` subdirectory of the output directory
+result HDF5 files. (This will be the `od2` subdirectory of the output directory
 from Phase 1 if you used the `typical_filtering_step.sh` script.)
 
 The script outputs a genetic similarity matrix to standard output by default. To
diff --git a/docs/tutorials/reads2tree/README.md b/docs/tutorials/reads2tree/README.md
index 1d800ff..ea1c3af 100644
--- a/docs/tutorials/reads2tree/README.md
+++ b/docs/tutorials/reads2tree/README.md
@@ -264,10 +264,10 @@ ls "$TUTORIAL_DIR/rna_clique_out/graph.pkl"
 
 If you want a tree, you can create one using RNA-clique and Biopython. The code
 below, also found in `docs/tutorials/reads2tree/make_tree.py`, computes the
-distance matrix from the `graph.pkl` and `od2/*.pkl` files and constructs a tree
-using the neighbor-joining algorithm. The tree is also rooted at its
-midpoint. The tree is saved to `nj_tree.tree`, and a visualization is saved to
-`nj_tree.svg` in the `rna_clique_out` directory.
+distance matrix from the `graph.pkl` and `od2/*.h5` (or `od2/*.pkl`) files and
+constructs a tree using the neighbor-joining algorithm. The tree is also rooted
+at its midpoint. The tree is saved to `nj_tree.tree`, and a visualization is
+saved to `nj_tree.svg` in the `rna_clique_out` directory.
 
 ```python
 --8<-- "docs/tutorials/reads2tree/make_tree.py"
diff --git a/docs/usage.md b/docs/usage.md
index 72d35a6..b588259 100644
--- a/docs/usage.md
+++ b/docs/usage.md
@@ -10,11 +10,11 @@ This script builds the gene matches graph from gene matches tables.
 
 ### Options
 
-| Short name | Long name | Description                       | Default | Required |
-|------------|-----------|-----------------------------------|---------|----------|
-| `-h`       | `--help`  | Print a help message and exit.    |         | No       |
-| `-i`       |           | Gene matches table pickles.       |         | Yes      |
-| `-o`       |           | Output gene matches graph pickle. |         | Yes      |
+| Short name | Long name | Description                              | Default | Required |
+|------------|-----------|------------------------------------------|---------|----------|
+| `-h`       | `--help`  | Print a help message and exit.           |         | No       |
+| `-i`       |           | Gene matches table HDF5 or pickle files. |         | Yes      |
+| `-o`       |           | Output gene matches graph pickle.        |         | Yes      |
 
 
 ## do\_filtering\_step.sh