diff --git a/docs/building/004_coexpression_network_cluster.md b/docs/building/004_coexpression_network_cluster.md
index 9299c51..5b204f8 100644
--- a/docs/building/004_coexpression_network_cluster.md
+++ b/docs/building/004_coexpression_network_cluster.md
@@ -1,7 +1,55 @@
-#### Adding coexpression network
+# Adding co-expression data and clusters
 
-*UNDER CONSTRUCTION*
+## Adding co-expression network
 
-#### Adding coexpression clusters
+Coexpression networks from [LSTrAP](https://github.molgen.mpg.de/proost/LSTrAP/) (Pearson Correlation based) can be directly imported
+and are converted to a rank-based network upon importing the file.
 
-*UNDER CONSTRUCTION*
\ No newline at end of file
+In the admin menu select 'Add' -> 'Add Co-Expression Network'. 
+
+![add_coexpression_network](../images/add_expression_network.png)
+
+**Select the species first** and enter a suitable **description**. Next select a 
+**Limit**, this specifies the maximum rank to keep. Lower values will retain fewer,
+but more reliable links where higher values result in a denser network.
+
+A **PCC-cutoff** can be combined with the rank based cutoff. Turn this to -2 to 
+disable the filter.
+
+For sparse networks, a gene's neighborhood can be extended from direct neighbors 
+only to include neighbors of neighbors (aka the second level). If you wish to do 
+so tick the checkbox **Enable Second Level Neighborhood?**. WARNING: Enabling this
+option for dense networks will significantly impact the performance of several of 
+CoNekT's features.
+
+Finally the full expression table as generated by LSTrAP should be selected. Click 
+**"Add Network"** to upload the file (note files can be large, this step can take a
+while) and import the data into the database. 
+
+## Adding co-expression clusters
+
+Co-expression clusters can either be imported or generated using a built-in 
+clustering algorithm (Heuristic Cluster Chiseling Algorithm). 
+
+### Importing MCL Clusters
+
+MCL based clusters (which are included in LSTrAP output) can be imported into LSTrAP.
+'Add' -> 'Coexpression Clusters' in the admin panel give you an interface where you 
+can **select an existing network**, add a **description** for the new clusters and 
+indicate a **minimal size** (to avoid including small clusters). Finally select the **MCL
+output** using **Select File** and click **Add Clusters**
+
+![adding cluster](../images/add_coexpression_clusters.png)
+
+### Building HCCA Clusters
+
+Importing MCL clusters can have the disadvantage that the imported network is
+filtered differently and inconsistencies emerge between clusters and networks. To
+remedy this, CoNekT allows admins to cluster the imported network using the HCCA.
+
+![building cluster](../images/build_coexpression_clusters.png)
+
+This is very straightforward, but can be time consuming. Make sure the webserver 
+doesn't have a timeout (or use the build in one). 
+
+Select a network, add a description for the clusters and click **Build Clusters**.
\ No newline at end of file
diff --git a/docs/images/add_coexpression_clusters.png b/docs/images/add_coexpression_clusters.png
new file mode 100644
index 0000000..7dd621a
Binary files /dev/null and b/docs/images/add_coexpression_clusters.png differ
diff --git a/docs/images/add_expression_network.png b/docs/images/add_expression_network.png
new file mode 100644
index 0000000..4da56d8
Binary files /dev/null and b/docs/images/add_expression_network.png differ
diff --git a/docs/images/add_expression_specificty.png b/docs/images/add_expression_specificty.png
new file mode 100644
index 0000000..cb7de20
Binary files /dev/null and b/docs/images/add_expression_specificty.png differ
diff --git a/docs/images/add_xrefs_genes.png b/docs/images/add_xrefs_genes.png
new file mode 100644
index 0000000..47874cd
Binary files /dev/null and b/docs/images/add_xrefs_genes.png differ
diff --git a/docs/images/build_coexpression_clusters.png b/docs/images/build_coexpression_clusters.png
new file mode 100644
index 0000000..88484bd
Binary files /dev/null and b/docs/images/build_coexpression_clusters.png differ