diff --git a/README.md b/README.md
index b6544e0de782c397ab19162641e2e05b0d2ef34f..02908a9a2220c5e9588a62f491633c1e9e73120b 100644
--- a/README.md
+++ b/README.md
@@ -9,11 +9,11 @@ The pipeline is basically written in perl.
 
 ## How does it work?
 It requires an important number of pre-computed datasets: 
-- A Core Genome and its Phylogenetic reconstructions for:
-    - The Phylogenetic clade of interest
-    - Two phylogenetic clades, outgroups of the clade of interest.
+- A Core Genome built with 3 clades and its Phylogenetic reconstruction:
+    - The clade of interest
+    - Two clades, outgroups of the clade of interest.
 
-*The core genomes for the three clades must be formated as HMMer database*
+Each core-gene alignment has to be divided by clade, so we can distinguish between the clade of interest and the outgroups. *The core genes for the three clades must be formated as HMMer database*
 
 - The Multiple sequence alignments of each Core-Gene
 - A list of reference sequences