diff --git a/ChangeLog b/ChangeLog
index 1099f64b..43d55d5c 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,8 @@
+Version 0.38.2
+  * Miscellanous bug files and updates to reference file handling.
+  * Updates to reference format documentation to describe how to get better IS prediction.
+  * Added experimental breseq SOFT-CLIPPING subcommand.
+
 Version 0.38.1
   * Fix for infinite loop triggered by help when terminal window was too narrow.
   
diff --git a/configure.ac b/configure.ac
index 4434759f..85d7e80e 100644
--- a/configure.ac
+++ b/configure.ac
@@ -19,7 +19,7 @@
 #                                               -*- Autoconf -*-
 # Process this file with autoconf to produce a configure script.
 AC_PREREQ([2.65])
-AC_INIT([breseq], [0.38.1], [jeffrey.e.barrick@gmail.com], [breseq], [http://barricklab.org/breseq])
+AC_INIT([breseq], [0.38.2], [jeffrey.e.barrick@gmail.com], [breseq], [http://barricklab.org/breseq])
 AC_CONFIG_AUX_DIR(aux_build)
 AC_CONFIG_MACRO_DIR([aux_build/m4])
 AC_CONFIG_HEADERS([aux_build/config.h])
diff --git a/src/c/breseq/settings.cpp b/src/c/breseq/settings.cpp
index db215951..e174949c 100644
--- a/src/c/breseq/settings.cpp
+++ b/src/c/breseq/settings.cpp
@@ -359,6 +359,7 @@ namespace breseq
 		options.addUsage("Utility Command Usage: breseq [command] options ...");
     options.addUsage("  Sequence Utility Commands: CONVERT-FASTQ, CONVERT-REFERENCE, GET-SEQUENCE");
     options.addUsage("  Breseq Post-Run Commands: BAM2ALN, BAM2COV, CL-TABULATE");
+    options.addUsage("  Other Commands: SOFT-CLIPPING");
     options.addUsage("");
     options.addUsage("For help using a utility command, type: breseq [command] "); 
     options.addUsage("");
diff --git a/src/doc/refseq_format.rst b/src/doc/refseq_format.rst
index a42cab97..19e08ee8 100644
--- a/src/doc/refseq_format.rst
+++ b/src/doc/refseq_format.rst
@@ -102,6 +102,37 @@ If multiple feature lines have identical accessions and types, then the location
 
 Pseudogenes are marked by adding ``Pseudo=true`` to the semicolon-delimited list of attributes at the end of the feature line line. Additionally, pseudogenes are reassigned a different feature type of ``fCDS``.
 
+Adding IS Element Annotations
+-------------------------------
+
+Many sequence files don't have IS elements annotated. To have |breseq| automatically predict IS elements as single events versus two JC evidence items that you have to figure out, we highly recommend adding these annotations. You can accomplish it using these steps:
+
+1. Install and run `ISEScan <https://github.com/xiezhq/ISEScan>`_ to generate a CSV file of IS predictions.
+
+  .. code-block:: bash
+
+    isescan.py --nthread 4 --seqfile reference.fasta --output output
+
+  .. note::
+
+    If you don't have a FASTA version of your reference, you can generate one using ```breseq CONVERT-REFERENCE``.
+
+  .. note::
+
+    The current version of ISEScan and its dependencies installed through Conda crashes on MacOSX due to a problem with FragGeneScan. If you really want to get it running on a Mac, you can install ISEScan via Conda and then install this fixed version inside of the same environment: `FragGeneScan with bug fix <https://github.com/barricklab/FragGeneScan>`_ .
+
+2. Merge these predictions into your reference file using ```breseq CONVERT-REFERENCE``.
+
+  .. code-block:: bash
+
+    breseq CONVERT-REFERENCE -f GENBANK -s output/reference.fasta.csv -o reference_with_IS.gbk reference.gbk
+
+  .. note::
+
+    You can also output as a GFF3 (substitute ``-f GFF3`` and ``-o reference_with_IS.gff``).
+
+3. Now run |breseq| with the updated reference file!
+
 Illegal Characters
 --------------------