diff --git a/scripts/README.md b/scripts/README.md
index 1332dc7..ffbd1a3 100644
--- a/scripts/README.md
+++ b/scripts/README.md
@@ -2,3 +2,10 @@
 
 This folder houses scripts that can be used to generate predicted mappings, typically
 through a lexical mapping workflow.
+
+Most of the lexical mappings in Biomappings were generated with a workflow that wraps Gilda and PyOBO.
+However, Biomappings is generic to any workflow that generates predictions, such as those
+coming from knowledge graph embedding models. More information can be found about the helper functions
+for writing your own prediction generation workflow can be found
+at https://biomappings.readthedocs.io/en/latest/usage.html. This also has a summary of the data types that
+correspond to rows in the mappings (`MappingTuple`) and predictions files (`PredictionTuple`).
diff --git a/src/biomappings/resources/__init__.py b/src/biomappings/resources/__init__.py
index 591c0bd..fb4dd97 100644
--- a/src/biomappings/resources/__init__.py
+++ b/src/biomappings/resources/__init__.py
@@ -152,7 +152,7 @@ class PredictionTuple(NamedTuple):
     the rough standard that closer to 1 is more confident and closer to 0 is less confident.
 
     Most of the lexical mappings already in Biomappings were generated with Gilda.
-    they were generated using Gilda. Depending on the script, the score therefore refers to either:
+    Depending on the script, the score therefore refers to either:
 
     1. The Gilda match score, inspired by https://aclanthology.org/W15-3801/. Section 5.2 of the
        `supplementary material for the Gilda paper <https://doi.org/10.1093/bioadv/vbac034>`_
@@ -160,7 +160,7 @@ class PredictionTuple(NamedTuple):
        https://github.com/biopragmatics/biomappings/blob/master/scripts/generate_agrovoc_mappings.py
        is an example that uses this variant.
     2. A high-level estimation of the precision of the scores generated by the given script.
-       For example, the CL-MeSH mappings were estimated to be 90% correct, so all of the mappings
+       For example, the CL-MeSH mappings were estimated to be 90% correct, so all the mappings
        generated by https://github.com/biopragmatics/biomappings/blob/master/scripts/generate_cl_mesh_mappings.py
        are marked with 0.9 as its score.