+ <li><strong>New dialog and faster and more
+ configurable Tree and PCA calculations</strong><br> Menu entries for
+ calculating PCA and different types of tree have been replaced by
+ a single <a href="calculations/calculations.html"><em>Calculations</em>
+ dialog box</a>. The underlying implementation for the PCA and tree
+ calculations have been made faster and more memory efficient.</li>
+ <li><strong>Extensible score models</strong><br />A new
+ framework has also been created for the score models used to
+ calculate distances between sequences and shade alignments. This
+ framework allows import of substitution matrices in NCBI and
+ AAIndex format.<br /> <strong>PCA Bug Fixes</strong>. Jalview's
+ implementation of PCA differed in its treatment of gaps and
+ non-standard residues. The BLOSUM62 matrix also included a typo
+ that affected results. See the <a
+ href="releases.html#2102scoremodelbugs">2.10.2 release note
+ about score model bugs</a> for details and how to reinstate legacy
+ behaviour.</li>
+ <li><strong>Update to JABAWS 2.2</strong><br />Jalview's
+ alignment, protein conservation analysis, and protein disorder and
+ RNA secondary structure prediction services are now provided by <a
+ href="http://www.compbio.dundee.ac.uk/jabaws">JABAWS 2.2</a>.
+ Several of the programs provided as JABAWS 2.2 services have been
+ updated, so their options and parameters have changed.</li>
+ <li><strong>URL linkouts to other bioinformatics
+ databases</strong><br />New preferences for <a
+ href="webServices/urllinks.html">opening web pages for
+ database cross-references</a> via the UK Elixir's EMBL-EBI's MIRIAM
+ database and identifiers.org services.</li>
+ <li><strong>Showing and hiding regions</strong> <br /> <a
+ href="menus/popupMenu.html#hideinserts">Hide insertions</a> in the
+ PopUp menu has changed its behaviour. Prior to 2.10.2, columns
+ were only shown or hidden according to gaps in the sequence under
+ the popup menu. Now, only columns that are gapped in all selected
+ sequences as well as the sequence under the popup menu are hidden,
+ and column visibility outside the selected region is left as is.
+ This makes it easy to filter insertions from the alignment view
+ (just select the region containing insertions to remove) without
+ affecting the rest of the hidden columns.</li>
+ <li><strong>Gap count - a.k.a. the Occupancy
+ Annotation Row</strong><br /> Another way to filter columns according to
+ the presence of gaps is to enable the <strong>Occupancy
+ Annotation</strong> row via Jalview's Preferences. This annotation row
+ shows a histogram of the number of aligned residues at each
+ column. The <a href="features/columnFilterByAnnotation.html">Select
+ By Annotation</a> dialog now also includes a percentage threshold
+ mode, to make it easy to filter alignments to show only those
+ columns with a particular fraction of aligned sequences.</li>
+ <li><strong>Recent search history for Find, PDBe and
+ Uniprot</strong><br />Easily repeat a previous search for <a
+ href="features/search.html#queryhistory">Find</a> and the free
+ text search system (for querying Uniprot and the PDBe).</li>
+ <li><strong>Improved Overview Window</strong><br />The <a
+ href="features/overview.html">alignment overview</a> is now easier
+ to use when working with alignments of more than 5000 rows and
+ columns, and features a new pop-up menu that allows hidden regions
+ to be excluded from the overview. It also works with CDS/Protein
+ alignments and MSA views in wrapped mode.</li>
+ <li><strong>3D Structure</strong><br />Jalview's communication
+ with UCSF Chimera has been made more robust, particularly when
+ working with many structures and long sequences. Regions in
+ structures that correspond to hidden regions in an alignment view
+ are now left un-coloured, making it easier to highlight specific
+ features in 3D. See below for <a href="#experimental">experimental
+ features for exchanging annotation between Chimera and Jalview.</a></li>