-<html>\r
-<head><title>Hidden Regions</title></head>\r
-<body>\r
-Now you see them, now you dont \r
-</body>\r
-</html>\r
+<html>
+<head><title>Hidden Regions</title></head>
+<body>
+<p><strong>Hidden Regions</strong> </p>
+<p>Use the keyboard key "H" to hide / reveal selected columns and sequences.
+ To hide / reveal only selected sequences, use "Shift H", to hide /
+ reveal only selected columns, use "Control H".</p>
+<p><strong><em>Hiding Sequences</em></strong><br>
+To hide selected sequences in an alignment, use the <strong>"View
+ -> Hide -> Selected Sequences"</strong> menu item or simply select <strong>"Hide
+ Sequences"</strong> from the Popup menu after a right click on the sequence Ids.
+</p>
+<p>Hidden sequences will not be used in any calculations, editing or web service
+ alignments performed on visible sequences. </p>
+<p><strong><em>Hidden Sequences Representatives</em></strong><br>
+A more advanced hide involves a right-mouse click on a sequence, then selecting
+ <strong>"SequenceID -> Represent Group with SequenceId"</strong>. Using this method
+ of hiding sequences, any edits performed on the visible group representative
+ will be propogated to all the sequences in that group. <br>
+ The hidden representative sequences will not be used in any
+ calculations or web service alignments (<em>nb. this may change in
+ the future</em>)
+</p>
+<p><strong><em>Hiding Columns</em></strong><br>
+To hide selected columns in an alignment, use the <strong>"View
+ -> Hide -> Selected Columns"</strong> menu item, or right click within a region
+ of selected columns in the scale above the alignment (only available in non
+ wrapped mode) and select <strong>"Hide Columns"</strong>. </p>
+<p>When an alignment view contains hidden columns, certain constraints
+apply:<ul><li>Editing the alignment is bound by the hidden columns, <em>i.e.</em> you cannot move residues
+ across a hidden column boundary.</li>
+<li><strong>Tree</strong>, <strong>pairwise alignment</strong> and <strong>PCA</strong> calculations will only be
+performed using the <em>visible</em> parts of the alignment.
+</li>
+<li><a
+ href="../webservices/msaclient.html">Multiple Sequence
+ Alignments</a> are performed locally on on each visible chunk of
+ the input, and concatenated with the hidden regions to form the
+ final result.</p>
+</li>
+</ul>
+<p><strong>Column Separability</strong><br>Calculations where hidden columns are
+ excluded, and a single analysis performed on the result, are termed <em>column-separable</em>.
+ The simple Tree and PCA calculations are column separable because essentially the same results
+ would be obtained if the excluded hidden columns were replaced by gaps as the input to the calculation.</p>
+ <p>Multiple Sequence alignment and secondary structure prediction are both non-column-separable, and
+ so the exclusion of hidden regions leads to only 'locally optimal' results - sometimes
+ different to that obtained when using the full alignment.</p>
+<p> </p>
+<p> </p>
+<p> </p>
+</body>
+</html>