diff --git a/README.md b/README.md
index e2257d7f7c0ed766fbbdee006d3e9d9501cc2b04..395107b6aa35f5adebac1d910ee023fe6c86aa01 100644
--- a/README.md
+++ b/README.md
@@ -1,15 +1,16 @@
 [![Build Status](https://travis-ci.org/hammerlab/mhcflurry.svg?branch=master)](https://travis-ci.org/hammerlab/mhcflurry) [![Coverage Status](https://coveralls.io/repos/github/hammerlab/mhcflurry/badge.svg?branch=master)](https://coveralls.io/github/hammerlab/mhcflurry?branch=master)
 
 # mhcflurry
-Open source peptide/MHC I binding affinity prediction. Competitive accuracy, fast,
-[documented](http://www.hammerlab.org/mhcflurry/).
+([MHC I](https://en.wikipedia.org/wiki/Major_histocompatibility_complex) ligand
+prediction package with competitive accuracy and a fast and [documented](http://www.hammerlab.org/mhcflurry/)
+implementation.
 
 MHCflurry supports Class I peptide/MHC binding affinity prediction using
 ensembles of allele-specific models. You can fit MHCflurry models to your own data
 or download models that we fit to data from
 [IEDB](http://www.iedb.org/home_v3.php) and [Kim 2014](http://bmcbioinformatics.biomedcentral.com/articles/10.1186/1471-2105-15-241).
 
-MHCflurry supports Python versions 2.7 and 3.4+. It uses the [keras](https://keras.io)
+MHCflurry runs on Python versions 2.7 and 3.4+. It uses the [keras](https://keras.io)
 neural network library via either the Tensorflow or Theano backends. GPUs may
 optionally be used for a generally modest speed improvement.