small improvements

Andreas Prlic · Andreas Prlic · commit a44d299e7f1d · 2013-09-19T13:57:11.000-07:00
diff --git a/structure/bioassembly.md b/structure/bioassembly.md
@@ -73,7 +73,7 @@ Let's load both representations of hemoglobin PDB ID [1HHO](http://www.rcsb.org/
     </tr>
 </table>
 
-As we can see, the two representations are quite different!
+As we can see, the two representations are quite different! When investigating protein interfaces, ligand binding and for many other applications, you always want to work with the biological assemblies!
 
 ## Further Reading
 
diff --git a/structure/firststeps.md b/structure/firststeps.md
@@ -30,8 +30,6 @@ BioJava  automatically downloads the PDB file for hemoglobin [4HHB](http://www.r
     </tr>
 </table>
 
-
-
 This demonstrates two things:
 
 + BioJava can automatically download and install files locally (more on this in Chapter 4)
@@ -43,6 +41,14 @@ If you already have a local PDB installation, you can configure where BioJava sh
     -DPDB_DIR=/wherever/you/want/
 </pre>
 
+## Memory consumption
+
+Talking about startup properties, it is also good to mention the fact that many PDB entries are large molecules and the default 64k memory allowance for Java applications is not sufficient in many cases.  BioJava contains several built-in caches which automatically adjust to the available memory. As such, the more memory you grant your Java applicaiton, the better it can utilize the caches and the better the performance will be. Change the maximum heap space of your Java VM with this startup parameter:
+
+<pre>
+    -Xmx1G
+</pre>
+
 ## A Quick 3D View
 
 If you have the *biojava-structure-gui* module installed, you can quickly visualise a [Structure](http://www.biojava.org/docs/api/org/biojava/bio/structure/Structure.html) via this:
