+\subsubsection{Saving Parameter Sets}
+When creating a custom parameter set, you will be asked for a file name to save
+it. The location of the file is recorded in the Jalview user preferences in the
+same way as a custom alignment colourscheme, so when Jalview is launched again,
+it will show your custom preset amongst the options available for running the
+JABA service.
+
+%
+% \exercise{Creating and using user defined presets}{\label{createandusepreseex}
+% \exstep{Import the file at
+% \textsf{http://www.jalview.org/tutorial/fdx\_unaligned.fa} into jalview.}
+% \exstep{Use the `{\slDiscover Database Ids}' function to recover the PDB cross
+% references for the sequences.}
+% \exstep{Align the sequences using the default ClustalW parameters.}
+% \exstep{Use the `{\sl Edit and run..}'
+% option to open the ClustalW parameters dialog box, and create a new preset using
+% the following settings:
+% \begin{list}{$\bullet$}{}
+% \item BLOSUM matrix (unchanged)
+% \item Gap Opening and End Gap penalties = 0.05
+% \item Gap Extension and Separation penalties = 0.05
+% \end{list}
+%
+% As you edit the parameters, buttons will appear on the dialog box
+% allowing you revert your changes or save your settings as a new parameter
+% set.
+%
+% Before you save your settings, remeber to give them a meaningful name by editing
+% the text box at the top of the dialog box.
+% }
+% \exstep{Repeat the alignment using your new parameter set by selecting it from
+% the {\sl ClustalW Presets menu}.}
+% \exstep{These sequences have PDB structures associated with them, so it is
+% possible to compare the quality of the alignments.
+%
+% Use the {\sl View all {\bf N}
+% structures} option to calculate the superposition of 1fdn on 1fxd for both
+% alignments (refer to section \ref{superposestructs} for instructions). Which
+% alignment gives the best RMSD ? }
+% \exstep{Apply the same alignment parameter settings to the example alignment
+% (available from \textsf{http://www.jalview.org/examples/uniref50.fa}).
+%
+% Are there differences ? If not, why not ?
+% }
+% }
+
+\section{Protein Alignment Conservation Analysis}
+\label{aacons}
+The {\sl Web Service $\Rightarrow$ Conservation} menu controls the computation
+of up to 17 different amino acid conservation measures for the current alignment
+view. The JABAWS AACon Alignment Conservation Calculation Service, which is used
+to calculate these scores, provides a variety of standard measures described by
+Valdar in 2002\footnote{Scoring residue conservation. Valdar (2002) {\sl
+Proteins: Structure, Function, and Genetics} {\bf 43} 227-241.} as well as an efficient implementation of the SMERFs
+score developed by Manning et al. in 2008.\footnote{SMERFS Score Manning et al. {\sl BMC
+Bioinformatics} 2008, {\bf 9} 51 \href{http://dx.doi.org/10.1186/1471-2105-9-51}{doi:10.1186/1471-2105-9-51}}
+
+\subsubsection{Enabling and Disabling AACon Calculations}
+When the AACon Calculation entry in the {\sl Web Services $\Rightarrow$
+Conservation} menu is ticked, AACon calculations will be performed every time
+the alignment is modified. Selecting the menu item will enable or disable
+automatic recalculation.
+
+\subsubsection{Configuring which AACon Calculations are Performed}
+The {\sl Web Services $\Rightarrow$ Conservation $\Rightarrow$ Change AACon
+Settings ...} menu entry will open a web services parameter dialog for the
+currently configured AACon server. Standard presets are provided for quick and
+more expensive conservation calculations, and parameters are also provided to
+change the way that SMERFS calculations are performed.
+AACon settings for an alignment are saved in Jalview projects along with the
+latest calculation results.
+
+\subsubsection{Changing the Server used for AACon Calculations}
+If you are working with alignments too large to analyse with the public JABAWS
+server, then you will most likely have already configured additional JABAWS
+servers. By default, Jalview will chose the first AACon service available from
+the list of JABAWS servers available. If available, you can switch to use
+another AACon service by selecting it from the {\sl Web Services $\Rightarrow$
+Conservation $\Rightarrow$ Switch Server} submenu.
+
+\chapter{Analysis of Alignments}
+\label{alignanalysis}
+Jalview provides support for sequence analysis in two ways. A number of
+analytical methods are `built-in', these are accessed from the {\sl Calculate}
+alignment window menu. Computationally intensive analyses are run outside
+Jalview {\sl via} web services - these are typically accessed {\sl via} the {\sl
+Web Service} menu, and described in chapter \ref{jvwebservices}.
+In this section, we describe the built-in analysis capabilities common to both
+the Jalview Desktop and the JalviewLite applet.
+
+\section{PCA}
+This calculation creates a spatial representation of the similarities within the
+current selection or the whole alignment if no selection has been made. After
+the calculation finishes, a 3D viewer displays each sequence as a point in
+3D `similarity space'. Sets of similar sequences tend to lie near each other in
+this space.
+Note: The calculation is computationally expensive, and may fail for very large
+sets of sequences - because the JVM has run out of memory. Memory issues, and
+how to overcome them, were discussed in Section \ref{memorylimits}.
+
+\subsubsection{What is PCA?}
+Principal components analysis is a technique for examining the structure of
+complex data sets. The components are a set of dimensions formed from the
+measured values in the data set, and the principal component is the one with the
+greatest magnitude, or length. The sets of measurements that differ the most
+should lie at either end of this principal axis, and the other axes correspond
+to less extreme patterns of variation in the data set.
+In this case, the components are generated by an eigenvector decomposition of
+the matrix formed from the sum of pairwise substitution scores at each aligned
+position between each pair of sequences. The basic method is described in the
+1995 paper by {\sl G. Casari, C. Sander} and {\sl A. Valencia} \footnote{{\sl
+Nature Structural Biology} (1995) {\bf 2}, 171-8.
+PMID: 7749921} and implemented at the SeqSpace server at the EBI.
+
+Jalview provides two different options for the PCA calculation. Protein PCAs are
+by default computed using BLOSUM 62 pairwise substitution scores, and nucleic
+acid alignment PCAs are computed using a score model based on the identity
+matrix that also treats Us and Ts as identical, to support analysis of both RNA
+and DNA alignments. The {\sl Change Parameters} menu also allows the calculation
+method to be toggled between SeqSpace and a variant calculation that is detailed
+in Jalview's built in documentation.\footnote{See
+\url{http://www.jalview.org/help/html/calculations/pca.html}.}
+
+\subsubsection{The PCA Viewer}
+
+PCA analysis can be launched from the {\sl Calculate $\Rightarrow$ Principal
+Component Analysis} menu option. {\bf PCA requires a selection containing at
+least 4 sequences}. A window opens containing the PCA tool (Figure \ref{PCA}).
+Each sequence is represented by a small square, coloured by the background
+colour of the sequence ID label. The axes can be rotated by clicking and
+dragging the left mouse button and zoomed using the $\uparrow$ and $\downarrow$
+keys or the scroll wheel of the mouse (if available). A tool tip appears if the
+cursor is placed over a sequence. Sequences can be selected by clicking on them.
+[CTRL]-Click can be used to select multiple sequences.
+
+Labels will be shown for each sequence by toggling the {\sl View $\Rightarrow$
+Show Labels} menu option, and the plot background colour changed {\sl via} the
+{\sl View $\Rightarrow$ Background Colour..} dialog box. A graphical
+representation of the PCA plot can be exported as an EPS or PNG image {\sl via}
+the {\sl File $\Rightarrow$ Save As $\Rightarrow$ \ldots } submenu.
+
+\exercise{Principal Component Analysis}
+{ \exstep{Load the alignment at
+\textsf{http://www.jalview.org/tutorial/alignment.fa}.}
+\exstep{Select the menu option {\sl Calculate $\Rightarrow$ Principle Component Analysis}.
+A new window will open. Move this window within the desktop so that the tree,
+alignment and PCA viewer windows are all visible.
+Try rotating the plot by clicking and dragging the mouse on the plot in the PCA window.
+Note that clicking on points in the plot will highlight them on the alignment. }
+\exstep{ Select {\sl Calculate $\Rightarrow$ Calculate Tree $\Rightarrow$
+Neighbour Joining Using BLOSUM62}. A new tree window will appear.
+Place the mouse cursor on the tree window so that the
+tree partition location will divide the alignment into a number of groups, each of a different colour.
+Note how the colour of the sequence ID label matches both the colour of
+the partitioned tree and the points in the PCA plot.} }
+
+\begin{figure}[hbtp]
+\begin{center}
+\includegraphics[width=2in]{images/PCA1.pdf}
+\includegraphics[width=3in]{images/PCA3.pdf}
+\caption{{\bf PCA Analysis.} }
+\label{PCA}
+\end{center}
+\end{figure}
+
+
+
+\subsubsection{PCA Data Export}
+Although the PCA viewer supports export of the current view, the plots produced
+are rarely suitable for direct publication. The PCA viewer's {\sl File} menu
+includes a number of options for exporting the PCA matrix and transformed points
+as comma separated value (CSV) files. These files can be imported by tools such
+as {\bf R} or {\bf gnuplot} in order to graph the data.
+
+\section{Trees}
+\label{trees}
+Jalview can calculate and display trees, providing interactive tree-based
+grouping of sequences though a tree viewer. All trees are calculated {\sl via}
+the {\sl Calculate $\Rightarrow$ Calculate Tree $\Rightarrow$ \ldots} submenu.
+Trees can be calculated from distance matrices determined from \% identity or
+aggregate BLOSUM 62 score using either {\sl Average Distance} (UPGMA) or {\sl
+Neighbour Joining} algorithms. The input data for a tree is either the selected
+region or the whole alignment, excluding any hidden regions.
+
+On calculating a tree, a new window opens (Figure \ref{trees1}) which contains
+the tree. Various display settings can be found in the tree window {\sl View}
+menu, including font, scaling and label display options, and the {\sl File
+$\Rightarrow$ Save As} submenu contains options for image and Newick file
+export. Newick format is a standard file format for trees which allows them to
+be exported to other programs. Jalview can also read in external trees in
+Newick format {\sl via} the {\sl File $\Rightarrow$ Load Associated Tree} menu
+option. Leaf names on imported trees will be matched to the associated alignment
+- unmatched leaves will still be displayed, and can be highlighted using the
+{\sl View $\Rightarrow$ Mark Unlinked Leaves} menu option.
+
+
+\begin{figure}
+\begin{center}
+\includegraphics[width=2.5in]{images/trees1.pdf}
+\includegraphics[width=2.5in]{images/trees2.pdf}
+\includegraphics[width=1.25in]{images/trees4.pdf}
+\caption{{\bf Calculating Trees} Jalview provides four built in models for calculating trees.
+Jalview can also load precalculated trees in Newick format (right).}
+\label{trees1}
+\end{center}
+\end{figure}
+
+
+Clicking on the tree brings up a cursor across the height of the tree. The
+sequences are automatically partitioned and coloured (Figure \ref{trees2}). To
+group them together, select the {\sl Calculate $\Rightarrow$ Sort $\Rightarrow$
+By Tree Order $\Rightarrow$ \ldots} alignment window menu option and choose the
+correct tree. The sequences will then be sorted according to the leaf order
+currently shown in the tree view. The coloured background to the sequence IDs
+can be removed with {\sl Select $\Rightarrow$ Undefine Groups} from the
+alignment window menu. Note that tree partitioning will also remove any groups
+and colourschemes on a view, so create a new view ([CTRL-T]) if you wish to
+preserve these.
+
+\begin{figure}
+\begin{center}
+\includegraphics[width=5in]{images/trees3.pdf}
+\caption{{\bf Interactive Trees} The tree level cutoff can be used to designate
+groups in Jalview.}
+\label{trees2}
+\end{center}
+\end{figure}
+
+%\subsubsection{Multiple Views and Input Data recovery from PCA and Tree Viewers}
+% move to ch. 3 ?
+%Both PCA and Tree viewers are linked analysis windows. This means that their selection and display are linked to a particular alignment, and control and reflect the selection state for a particular view.
+
+\subsubsection{Recovering input Data for a Tree or PCA Plot Calculation}
+\parbox[c]{5in}{
+The {\sl File $\Rightarrow$ Input Data } option will open a new alignment window containing the original data used to calculate the tree or PCA plot (if available). This function is useful when a tree has been created and then the alignment subsequently changed.
+}
+\parbox[c]{1.25in}{\centerline{\includegraphics[width=1.25in]{images/pca_fmenu.pdf}
+}}
+
+\subsubsection{Changing the associated View for a Tree or PCA Viewer}
+\parbox[c]{4in}{
+The {\sl View $\Rightarrow$ Associated Nodes With $\Rightarrow$ .. } submenu is shown when the viewer is associated with an alignment that is involved in multiple views. Selecting a different view does not affect the tree or PCA data, but will change the colouring and display of selected sequences in the display according to the colouring and selection state of the newly associated view.
+} \parbox[c]{3in}{\centerline{
+\includegraphics[width=2.5in]{images/pca_vmenu.pdf} }}
+
+\subsection{Tree Based Conservation Analysis}
+\label{treeconsanaly}
+
+Trees reflect the pattern of global sequence similarity exhibited by the
+alignment, or region within the alignment, that was used for their calculation.
+The Jalview tree viewer enables sequences to be partitioned into groups based
+on the tree. This is done by clicking within the tree viewer window. Once subdivided, the
+conservation between and within groups can be visually compared in order to
+better understand the pattern of similarity revealed by the tree and the
+variation within the clades partitioned by the grouping. The conservation based
+colourschemes and the group associated conservation and consensus annotation
+(enabled using the alignment window's {\sl View $\Rightarrow$ Autocalculated
+Annotation $\Rightarrow$ Group Conservation} and {\sl Group Consensus} options)
+can help when working with larger alignments.
+
+\exercise{Trees}
+{Ensure that you have at least 1G memory available in Jalview
+(Either start with this link:
+\href{http://www.jalview.org/services/launchApp?jvm-max-heap=1G}{http://www.jalview.org/services/launchApp?jvm-max-heap=1G},
+or in the Development section of the Jalview web site
+(\href{http://www.jalview.org/development/development-builds}{http://www.jalview.org/development/development-builds})
+in the ``latest official build'' row in the table, go to the
+``Webstart'' column, click on ``G2''.)
+
+\exstep{Open the alignment at \textsf{http://www.jalview.org/tutorial/alignment.fa}.
+Select {\sl Calculate $\Rightarrow$ Calculate Tree $\Rightarrow$ Neighbour Joining Using BLOSUM62}. A new tree window will appear.}
+\exstep{Click on the tree window, a cursor will appear. Note that placing
+this cursor divides the tree into a number of groups by colour.
+Place the cursor to give about 4 groups.}
+\exstep{In the alignment window, select
+{\sl Calculate $\Rightarrow$ Sort $\Rightarrow$ By Tree Order $\Rightarrow$ Neighbour Joining Tree using BLOSUM62 from ... }. The sequences are
+reordered to match the order in the tree and groups are formed implicitly.
+Alternatively in the tree window, select {\sl View $\Rightarrow$ Sort Alignment
+by Tree}.} \exstep{Select {\sl Calculate $\Rightarrow$ Calculate Tree
+$\Rightarrow$ Neighbour Joining Using \% Identity}. A new tree window will appear. The group colouring makes it easy to see the differences between the two trees, calculated using
+ different methods.} \exstep{Select from sequence 2 column 60 to sequence 12 column 123. Select {\sl Calculate $\Rightarrow$ Calculate Tree $\Rightarrow$ Neighbour Joining Using BLOSUM62}. A new tree window will appear. It can be seen that the tree contains 11 sequences. It has been coloured according to the already selected groups from the first tree and is calculated purely from the residues in the selection.
+
+Comparing the location of individual sequences between the three trees illustrates the importance of selecting appropriate regions of the
+alignment for the calculation of trees. }}
+
+\exercise{Pad Gaps in an Alignment}{
+\exstep{Open the alignment at \textsf{http://www.jalview.org/tutorial/alignment.fa}. In alignment window, ensure that the {\sl Edit $\Rightarrow$
+Pad Gaps } option is {\sl not} ticked, and insert one gap anywhere in the
+alignment.}
+\exstep{Select {\sl Calculate $\Rightarrow$ Calculate Tree $\Rightarrow$
+Neighbour Joining Using BLOSUM62}.
+
+A warning dialog box {\bf ``Sequences not aligned'' } appears because the sequences input to the tree calculation are of different lengths. }
+
+\exstep{Select {\sl Edit $\Rightarrow$ tick Pad Gaps } and perform the
+tree calculation again. This time a new tree should appear - because padding
+gaps ensures all the sequences are the same length after editing.
+
+{\sl Pad Gaps } option
+can be set in Preferences using
+{\sl Tool $\Rightarrow$ Preference $\Rightarrow$ Editing}. }
+
+}
+
+\exercise{Tree Based Conservation Analysis}{
+\label{consanalyexerc}
+\exstep{Load the PF03460 PFAM seed alignment using the sequence fetcher.
+Select {\sl Colour $\Rightarrow$ Taylor $\Rightarrow$ By Conservation}
+, and set {\sl Conservation } shading threshold at around 20. }
+\exstep{Build a Neighbour joining tree using Select {\sl Calculate $\Rightarrow$ Calculate Tree $\Rightarrow$
+Neighbour Joining Using BLOSUM62}.}
+\exstep{Use the mouse cursor to select a point on the tree to partition the
+alignment into several sections.}
+\exstep {Select {\sl View $\Rightarrow$ Sort Alignment By Tree} option in the
+tree window to re-order the sequences in the alignment using the calculated
+tree.
+Examine the variation in colouring between different groups of sequences in the alignment
+window.}
+
+\exstep {You may find it easier to browse the alignment if you first uncheck the
+{\sl Annotations $\Rightarrow$ Show Annotations} option, and open the Overview Window
+within the View menu to aid navigation.}
+
+\exstep{Try changing the colourscheme of the residues in the alignment to
+BLOSUM62 (whilst ensuring that {\sl Apply Colour to All Groups} is selected).}
+{\sl Note: You may want to save the alignment and tree as a project file, since it is used in the next few exercises. } }
+
+
+\subsection{Redundancy Removal}
+
+The redundancy removal dialog box is opened using the {\sl Edit $\Rightarrow$ Remove Redundancy\ldots} option in the alignment menu. As its menu option placement suggests, this is actually an alignment editing function, but it is convenient to describe it here. The redundancy removal dialog box presents a percentage identity slider which sets the redundancy threshold. Aligned sequences which exhibit a percentage identity greater than the current threshold are highlighted in black. The [Remove] button can then be used to delete these sequences from the alignment as an edit operation\footnote{Which can usually be undone. A future version of Jalview may allow redundant sequences to be hidden, or represented by a chosen sequence, rather than deleted.}.
+\begin{figure}
+\begin{center}
+\includegraphics[width=5.5in]{images/redundancy.pdf}
+\end{center}
+\label{removeredundancydialog}
+\caption{The Redundancy Removal dialog box opened from the edit menu. Sequences that exceed the current percentage identity threshold and are to be removed are highlighted in black.}
+\end{figure}
+
+
+\subsection{Subdividing the Alignment According to Specific Mutations}
+
+It is often necessary to explore variations in an alignment that may correlate
+with mutations observed in a particular region; for example, sites exhibiting
+single nucleotide polymorphism, or residues involved in substrate recognition in
+an enzyme. One way to do this would be to calculate a tree using the specific
+region, and subdivide it in order to partition the alignment.
+However, calculating a tree can be slow for large alignments, and the tree may
+be difficult to partition when complex mutation patterns are being analysed. The
+{\sl Select $\Rightarrow$ Make groups for selection } function was introduced to
+make this kind of analysis easier. When selected, it will use the characters in
+the currently selected region to subdivide the alignment. For example, if a
+single column is selected, then the alignment (or each group defined on the
+alignment) will be divided into groups based on the residue or nucleotide found
+at that position. These new groups are annotated with the characters in the
+selected region, and Jalview's group based conservation analysis annotation and
+colourschemes can then be used to reveal any associated pattern of sequence
+variation across the whole alignment.
+
+
+% These annotations can be hidden and deleted via the context menu linked to the
+% annotation row; but they are only created on loading an alignment. If they are
+% deleted then the alignment should be saved and then reloaded to restore them.
+% Jalview provides a toggle to autocalculate a consensus sequence upon editing.
+% This is normally selected by default, but can be turned off for large alignments {\sl via} the {\sl Calculate $\Rightarrow$ Autocalculate
+% Consensus} menu option if the interface is too slow.
+
+% \subsubsection{Group Associated Annotation}
+% \label{groupassocannotation}
+% Group associated consensus and conservation annotation rows reflect the
+% sequence variation within a particular group. Their calculation is enabled
+% by selecting the {\sl Group Conservation} or {\sl Group Consensus} options in
+% the {\sl Annotation $\Rightarrow$ Autocalculated Annotation } submenu of the
+% alignment window.
+
+% \subsubsection{Alignment and Group Sequence Logos}
+% \label{seqlogos}
+
+% The consensus annotation row that is shown below the alignment can be overlaid
+% with a sequence logo that reflects the symbol distribution at each column of
+% the alignment. Right click on the Consensus annotation row and select the {\sl
+% Show Logo} option to display the Consensus profile for the group or alignment.
+% Sequence logos can be enabled by default for all new alignments {\sl via} the
+% Visual tab in the Jalview desktop's preferences dialog box.
+
+\section{Pairwise Alignments}
+Jalview can calculate optimal pairwise alignments between arbitrary
+sequences {\sl via} the {\sl Calculate $\Rightarrow$ Pairwise Alignments\ldots} menu option.
+Global alignments of all pairwise combinations of the selected sequences are performed and the results returned in a text box.
+
+
+
+\exercise{Remove Redundant Sequences}{
+
+\exstep{Using the alignment generated in the previous exercise (exercise
+\ref{consanalyexerc}).
+In the alignment window, you may need to deselect groups using Esc key.}
+
+\exstep{In the {\sl Edit} menu select {\sl Remove Redundancy} to open the
+Redundancy threshold selection dialog. Adjust the redundancy threshold value, start
+at 50 and increase the value to 65. Sequences selected will change colour in the Sequence ID panel. Select ``Remove'' to
+remove the sequences that are more than 65\% similar under this alignment.}
+
+\exstep{Select the Tree viewer's {\sl View $\Rightarrow$ Mark Unlinked Leaves} option, and note that the removed sequences are now prefixed with a * in the tree view.}
+\exstep{Use the [Undo] button in the Redundancy threshold selection dialog box
+to recover the sequences. Note that the * symbols disappear from the tree display.}
+\exstep{Experiment with the redundancy removal and observe the relationship between the percentage identity threshold and the pattern of unlinked nodes in the tree display.}
+}
+
+\exercise{Group Conservation Analysis}{
+\exstep{Re-use or recreate the alignment and tree which you worked with in the
+tree based conservation analysis exercise (exercise \ref{consanalyexerc})}
+\exstep{In the {\sl View} menu in the alignment window, select {\sl New View} to
+create a new view. Ensure the annotation panel is displayed ({\sl Show annotation} in {\sl Annotations} menu). Enable the
+display of {\sl Group Consensus} option by checking {\sl Group Consensus} in the {\sl Annotation $\Rightarrow$
+Autocalculated Annotation } submenu in the sequence alignment window.}
+\exstep{Displaying the sequence
+logos will make it easier to see the different residue populations within each
+group. Activate the logo by right clicking on the Consensus annotation row to
+open the context menu and select the {\sl Show Logo} option.}
+\exstep{In the column alignment ruler, select a column exhibiting about 50\%
+conservation that lies within the central conserved region of the alignment.
+(Column 74 is used in \href{https://youtu.be/m-PjynicXRg}{the Tree video}.}
+\exstep{Subdivide the alignment
+according to this selection using {\sl Select $\Rightarrow$ Make groups for selection}.}
+\exstep{Re-order the alignment according to the new groups that have been
+defined by selecting {\sl Calculate $\Rightarrow$ Sort $\Rightarrow$
+By Group}.
+
+Click on the group annotation row IDs to select groups exhibiting a
+specific mutation.}
+\exstep{Select another column exhibiting about 50\% conservation
+overall, and subdivide the alignment further. Note that the new groups
+inherit the names of the original groups, allowing you to identify the
+combination of mutations that resulted in the subdivision.
+}
+\exstep{Clear the groups, and try to subdivide the alignment using two
+non-adjacent columns.
+
+{\sl Hint: You may need to hide the intervening columns before you can select
+both of the columns that you wish to use to subdivide the alignment.}}
+\exstep{Switch back to the original view, and experiment with subdividing
+the tree groups made in the previous exercise.}
+}
+
+\begin{figure}[]
+\begin{center}
+\includegraphics[width=4in]{images/pairwise.pdf}
+\caption{{\bf Pairwise alignment of sequences.} Pairwise alignments of three selected sequences are shown in a textbox.}
+\label{pairwise}
+\end{center}
+\end{figure}
+
+
+\chapter{Working with 3D structures}
+\label{3Dstructure}
+Jalview can interactively view 3D structure using Jmol or Chimera. Setting in
+the Structure window within Preferences determine whether Jmol or Chimera is
+the default choice of structure viewer.
+
+% To review, Chapter \ref{featannot} describes the mechanisms provided by
+% Jalview for interactive creation of sequence and alignment annotation, and how
+% they can be displayed, imported and exported and used to reorder the alignment. Chapter
+% \ref{featuresfromdb} discusses the retrieval of database references and
+% establishment of sequence coordinate systems for the retrieval and display of
+% features from databases and DAS annotation services.
+% Chapter \ref{msaservices} describes how to use the range of multiple alignment
+% programs provided by JABAWS, and Chapter \ref{aacons} introduces JABAWS AACon
+% service for protein multiple alignment conservation analysis.
+% In Chapter \ref{alignanalysis}, you will find
+% descriptions and exercises on building and displaying trees, PCA analysis,
+% alignment redundancy removal, pairwise alignments and alignment conservation
+% analysis.
+% Chapter \ref{wkwithstructure} introduces the structure visualization
+% capabilities of Jalview.
+% Chapter \ref{protsspredservices} explains how to perform protein secondary
+% structure predictions with JPred, and JABAWS protein disorder prediction
+% services are introduced in Chapter \ref{protdisorderpred}.
+% Chapter \ref{workingwithnuc} describes functions and visualization techniques
+% relevant to working with nucleotide sequences, coding region annotation and nucleotide
+% sequence alignments.
+% Chapter \ref{jvwebservices} introduces the various web based services
+% available to Jalview users, and Chapter \ref{jabaservices} explains how to
+% configure the Jalview Desktop for access to new JABAWS servers.
+
+
+% and Section \ref{workingwithrna} covers the visualization,
+% editing and analysis of RNA secondary structure.
+
+\section{Working with Structures}
+\label{wkwithstructure}
+Jalview facilitates the use of protein structures for the analysis of alignments
+by providing a linked view of structures associated with sequences in
+the alignment. The Java based molecular viewing program Jmol\footnote{See the
+Jmol homepage \url{http://www.jmol.org} for more information.} has been
+incorporated\footnote{Earlier versions of Jalview included MCView - a simple
+main chain structure viewer. Structures are visualized as an alpha carbon trace
+and can be viewed, rotated and coloured in a structure viewer and the results
+interpreted on a sequence alignment.} which enables sophisticated molecular
+visualizations to be prepared and investigated alongside an analysis of
+associated sequences.
+PDB format files can be imported directly or structures can be retrieved from
+the European Protein Databank (PDBe) using the Sequence Fetcher (see
+\ref{fetchseq}).
+
+\subsection{Automatic Association of PDB Structures with Sequences}
+Jalview can automatically determine which structures are associated with a
+sequence in a number of ways.
+\subsubsection{Discovery of PDB IDs from Sequence Database Cross-references}
+If a sequence has an ID from a public database that contains cross-references to
+the PDB, such as Uniprot. Right-click on any sequence ID and select {\sl Structure $\Rightarrow$
+Associate Structure with Sequence $\Rightarrow$ Discover PDB IDs } from the context menu (Figure \ref{auto}). Jalview will attempt to associate the
+sequence with a Uniprot sequence and from there discover any associated PDB
+structures. This takes a few seconds and applies to all sequences in the
+alignment which have valid Uniprot IDs. On moving the cursor over the sequence
+ID the tool tip\footnote{Tip: The sequence ID tooltip can often become large for
+heavily cross referenced sequence IDs. Use the {\sl View $\Rightarrow$ Sequence
+ID Tooltip $\Rightarrow$ } submenu to disable the display of database cross
+references or non-positional features. } now shows the Uniprot ID and any
+associated PDB structures.
+
+\begin{figure}[htbp]
+\begin{center}
+%TODO fix formatting
+\parbox{1.5in}{
+{\centering
+\begin{center}
+\includegraphics[width=1.5in]{images/auto1.pdf}
+\end{center}}
+} \parbox{3.25in}{
+{\centering
+\begin{center}
+\includegraphics[scale=0.5]{images/auto2.pdf}
+\end{center}
+}
+} \parbox{1.5in}{
+{\centering
+\begin{center}
+\includegraphics[width=1.5in]{images/auto3.pdf}
+\end{center}
+}
+}
+
+\caption{{\bf Automatic PDB ID discovery.} The tooltip (left) indicates that no PDB structure has been associated with the sequence.
+After PDB ID discovery (center) the tool tip now indicates the Uniprot ID and
+any associated PDB structures (right).}
+\label{auto}
+\end{center}
+\end{figure}
+
+\subsubsection{Drag-and-Drop Association of PDB Files with Sequences by Filename
+Match}
+\label{multipdbfileassoc}
+If one or more PDB files stored on your computer are dragged from their location
+on the file browser onto an alignment window, Jalview will search the alignment
+for sequences with IDs that match any of the files dropped onto the alignment.
+If it discovers matches, a dialog like the one in Figure
+\ref{multipdbfileassocfig} is shown, giving the option of creating associations
+for the matches.
+
+If no associations are made, then sequences extracted
+from the structure will be simply added to the alignment. However, if only
+some of the PDB files are associated, Jalview will raise another dialog box
+giving you the option to add any remaining sequences from the PDB structure files not present in
+the alignment. This allows you to easily decorate sequences in a newly imported
+alignment with any corresponding structures you've already collected in a directory
+accessible from your computer.\footnote{We plan to extend this facility in
+future so Jalview will automatically search for PDB files matching your
+sequence within a local directory. Check out
+\href{http://issues.jalview.org/browse/JAL-801}{Jalview issue 801}}
+
+% there is no mention of the other footnote (#3) that appears saying: Tip: The sequence ID tooltip can often become large for heavily cross-referenced sequence IDs. Use the ...
+% JBP: yes there is - under 'Discovery of ' subsection.
+\begin{figure}[htbp]
+\begin{center}
+\includegraphics[]{images/pdbdragdropassoc.pdf}
+
+\caption{{\bf Associating PDB files with sequences by drag-and-drop.} Dragging
+PDB files onto an alignment of sequences with names matching the dragged files
+names (A), results in a dialog box (B) that gives the option to associate each
+file with any sequences with matching IDs. }
+\label{multipdbfileassocfig}
+\end{center}
+\end{figure}
+
+
+\subsection{Viewing Structures}
+\label{viewAllStructures}
+The structure viewer can be launched in two ways from the sequence ID context
+menu. To view a particular structure associated with a sequence in the
+alignment, simply select it from popup menu's associated structures submenu in
+{\sl Structure $\Rightarrow$ View Structure $\Rightarrow$ $<$PDB ID$>$}. The
+second way is most useful if you want to view all structural data available for
+a set of sequences in an alignment. If any of the {\bf currently selected}
+sequences have structures associated, the {\sl Structure } submenu of the
+sequence ID popup menu will include an option to {\sl View {\bf N}
+structures}. Selecting this option will open a new structure view containing
+the associated structures superposed according to the alignment.
+
+In both cases, each structure to be displayed will be downloaded or loaded from
+the local file system, and shown as a ribbon diagram coloured according to the
+associated sequence in the current alignment view (Figure \ref{structure}
+(right)). The structure can be rotated by clicking and dragging in the structure
+window. The structure can be zoomed using the mouse scroll wheel or by
+[SHIFT]-dragging the structure.
+Moving the mouse cursor over a sequence to which the structure is linked in the
+alignment view highlights the respective residue's sidechain atoms. The
+sidechain highlight may be obscured by other parts of the molecule. Similarly,
+moving the cursor over the structure shows a tooltip and highlights the
+corresponding residue in the alignment. Clicking the alpha carbon or phosphorous
+backbone atom will toggle the highlight and residue label on and off. Often, the
+position highlighted in the sequence may not be in the visible portion of the
+current alignment view and the sliders will scroll automatically to show the
+position. If the alignment window's {\sl View $\Rightarrow$ Automatic Scrolling
+} option is not selected, however, then the automatic adjustment will be
+disabled for the current view.
+
+\begin{figure}[htbp]
+\begin{center}
+\parbox{3in}{
+{\centering
+\begin{center}
+\includegraphics[scale=0.5]{images/structure1.pdf}
+\end{center}
+}
+}
+\parbox{3.2in}{
+{\centering
+\begin{center}
+\includegraphics[width=3in]{images/structure2.pdf}
+\end{center}
+}
+}
+\caption{{\bf Structure visualization} The structure viewer is launched from the sequence ID context menu (left) and allows the structure to be visualized using the embedded Jmol molecular viewer (right). }
+\label{structure}
+\end{center}
+\end{figure}
+
+\subsection{Customising Structure Display}
+
+Structure display can be modified using the {\sl Colour} and {\sl View} menus
+in the structure viewer. The background colour can be modified by selecting the
+{\sl Colours $\Rightarrow$ Background Colour\ldots} option.
+
+By default, the structure will be coloured in the same way as the associated
+sequence(s) in the alignment view from which it was launched. The structure can
+be coloured independently of the sequence by selecting an appropriate colour
+scheme from the {\sl Colours} menu. It can be coloured according to the
+alignment using the {\sl Colours $\Rightarrow$ By Sequence } option. The image
+in the structure viewer can be saved as an EPS or PNG with the {\sl File
+$\Rightarrow$ Save As $\Rightarrow$ \ldots} submenu, which also allows the raw
+data to be saved as PDB format. The mapping between the structure and the
+sequence (How well and which parts of the structure relate to the sequence) can
+be viewed with the {\sl File $\Rightarrow$ View Mapping} menu option.
+
+\subsubsection{Using the Jmol Visualization Interface }
+
+Jmol has a comprehensive set of selection and visualization functions that are
+accessed from the Jmol popup menu (by right-clicking in the Jmol window or by
+clicking the Jmol logo). Molecule colour and rendering style can be manipulated,
+and distance measurements and molecular surfaces can be added to the view. It
+also has its own ``Rasmol\footnote{See \url{http://www.rasmol.org}}-like'' scripting
+language, which is described elsewhere\footnote{Jmol Wiki:
+\url{http://wiki.jmol.org/index.php/Scripting}
+
+Jmol Scripting reference:
+\url{http://www.stolaf.edu/academics/chemapps/jmol/docs/}}. Jalview utilises the
+scripting language to interact with Jmol and to store the state of a Jmol
+visualization within Jalview archives, in addition to the PDB data file
+originally loaded or retrieved by Jalview. To access the Jmol scripting
+environment directly, use the {\sl Jmol $\Rightarrow$ Console} menu option.
+
+If you would prefer to use Jmol to manage structure colours, then select the
+{\sl Colours $\Rightarrow$ Colour with Jmol} option. This will disable any
+automatic application of colour schemes when new structure data is added, or
+when associated alignment views are modified.
+
+\exercise{Viewing Structures in Jmol viewer}{\label{viewingstructex}
+\exstep{Load the alignment at
+\textsf{http://www.jalview.org/examples/exampleFile.jar}.}
+\exstep{Right-click on the
+sequence ID label of {\sl FER1\_SPIOL} to open
+the context menu. Select {\sl 3D Structure}, this
+opens a Structure Chooser window, select { \sl 1A70} and click {\sl OK}.
+
+{\sl Note: the Structure Chooser interface
+provides a smart technique for selecting PDB structures by queryingthe meta-data
+of structures. Extra information can be including in this window by checking boxes
+in the ``Configure Displayed Columns'' tab}.
+% JBP Note: Bug JAL-1238 needs to be fixed ASAP
+}
+\exstep{By default the Jmol
+structure viewer opens in the Jalview desktop. Rotate the molecule by clicking
+and dragging in the structure viewing box.
+Zoom with the mouse scroll wheel. }
+\exstep{Roll the
+mouse cursor along the {\sl FER1\_SPIOL} sequence in the alignment.
+Note that if a residue in the sequence maps to one in the structure, a label
+will appear next to that residue in the structure viewer.}
+\exstep{Move the mouse over the structure. In the Jmol viewer, placing the mouse over a
+part of the structure will bring up a tool tip indicating the name and number of that residue.
+In the alignment window, the corresponding residue in the sequence is
+highlighted in black.}
+\exstep{Clicking the alpha carbon toggles the highlight and residue label on and
+off.
+Try this by clicking on a set of three or four adjacent residues so that the labels are persistent, then finding where they are in the sequence. }
+\exstep{In the structure viewer menu, select {\sl Colours $\Rightarrow$ Background
+Colour\ldots} and choose a suitable colour.
+Press {\sl OK} to apply this.}
+\exstep{Select {\sl File $\Rightarrow$ Save As $\Rightarrow$ PNG} and save the
+image. On your computer, view this with a suitable program. }
+\exstep{Select
+{\sl File $\Rightarrow$ View Mapping} from the structure viewer menu.
+A new window opens showing the residue by residue alignment between the sequence and the structure.}
+\exstep{Select {\sl File $\Rightarrow$ Save $\Rightarrow$ PDB file} and choose a new filename to save the PDB file.
+Once the file is saved, open the location in your file browser (or explorer window) and drag the PDB file that you just saved on
+to the Jalview desktop (or load it from the {\sl Jalview Desktop $\Rightarrow$ Input Alignment $\Rightarrow$ From File } menu).
+Verify that you can open and view the associated structure from the sequence ID
+pop-up menu's {\sl 3D Structure } submenu in the new alignment window.}
+
+\exstep{In the Jmol window, right click on the structure window and explore the
+menu options. Try to change the style of molecular display - for example by
+using the {\sl Jmol $\Rightarrow$ Select (n) $\Rightarrow$ All} command (where {\sl n} is the number of residues selected), and then the {\sl Jmol $\Rightarrow$ Style $\Rightarrow$ Scheme $\Rightarrow$ Ball and Stick} command.}
+\exstep{In the alignment window, use the {\sl File $\Rightarrow$ Save As .. }
+function to save the alignment as a Jalview Project. Now close the alignment and the structure view, and load the project file you just saved.
+Verify that the Jmol display is as it was when you just saved the file.}}
+
+\exercise{Setting Chimera as the default 3D Structure Viewer}{\label{viewingchimera}
+Jalview supports molecular structure
+visualization using both Jmol and Chimera 3D viewers. Jmol is the default
+viewer, however Chimera can be set up as the default choice from Preferences.
+
+\exstep{First, Chimera must be downloaded and installed on the computer.
+Chimera program is available on the UCSF web site \textsf{https://www.cgl.ucsf.edu/chimera/download.html}.}
+\exstep{In the desktop menu, select {\sl Tool $\Rightarrow$ Preferences}. In
+the ``{\sl
+Structure}'' tab set {\sl Default structure viewer} as {\sl
+Chimera}; then click {\sl OK}.}
+\exstep{Close the Jalview program, from the
+{\sl Desktop menu} select {\sl Jalview $\Rightarrow$ Quit Jalview}. Then reopen
+Jalview, Chimera should open as the default viewer.}
+{\sl Note: The Jmol structure viewer sits within the Jalview desktop; however
+the Chimera structure viewer sits outside the Jalview desktop and a Chimera
+view window sits inside the Jalview desktop.} }
+
+\subsection{Superimposing Structures}
+\label{superposestructs}
+Many comparative biomolecular analysis investigations aim to determine if the
+biochemical properties of a given molecule are significantly different to its
+homologues. When structure data is available, comparing the shapes of molecules
+by superimposing them enables substructure that may impart different behaviour
+to be quickly identified. The identification of optimal 3D superposition
+involves aligning 3D data rather than sequence symbols, but the result can
+still be represented as a sequence alignment, where columns indicate positions
+in each molecule that should be superposed to recreate the optimal 3D alignment.
+
+Jalview can employ Jmol's 3D fitting routines\footnote{See
+\href{http://chemapps.stolaf.edu/jmol/docs/?ver=12.2$\#$compare}{http://chemapps.stolaf.edu/jmol/docs/?ver=12.2$\#$compare}
+for more information.} to recreate 3D structure superpositions based on the
+correspondences defined by one or more sequence alignments involving structures shown in the Jmol display. Superposition based on the currently displayed alignment view happens automatically if a
+structure is added to an existing Jmol display using the {\sl Structure
+$\Rightarrow$ View PDB Structure $\Rightarrow$ ..}. A new Jmol view containing
+superposed structures can also be created using the {\sl Structure
+$\Rightarrow$ View all {\bf N} PDB Structures} option (when {\bf {\sl N}}
+$>$ 1) if the current selection contains two or more sequences with associated
+structures.
+
+\subsubsection{Obtaining the RMSD for a Superposition}
+The RMSD (Root Mean Square Deviation) is a measure of how similar the structures
+are when they are superimposed. Figure \ref{mstrucsuperposition} shows a
+superposition created during the course of Exercise \ref{superpositionex}. The
+parts of each molecule used to construct the superposition are rendered using
+the cartoon style, with other parts of the molecule drawn in wireframe. The Jmol
+console, which has been opened after the superposition was performed, shows the
+RMSD report for the superposition.
+Full information about the superposition is also outputted to the Jalview
+console.\footnote{The Jalview Java Console is opened from {\sl Tools
+$\Rightarrow$ Java Console} option in the Desktop's menu bar} This output also
+includes the precise atom pairs used to superpose structures.
+
+\subsubsection{Choosing which part of the Alignment is used for Structural
+Superposition} Jalview uses the visible part of each alignment view to define
+which parts of each molecule are to be superimposed. Hiding a column in a view
+used for superposition will remove that correspondence from the set, and will
+exclude it from the superposition and RMSD calculation.
+This allows the selection of specific parts of the alignment to be used for
+superposition. Only columns that define a complete set of correspondences for
+all structures will be used for structural superposition, and as a consequence,
+the RMSD values generated for each pair of structures superimposed can be
+directly compared.
+
+In order to recompute a superposition after changing a view or editing the
+alignment, select the {\sl Jmol $\Rightarrow$ Align sequences } menu option. The {\sl
+Jmol $\Rightarrow$ Superpose with ..} submenu allows you to choose which of the
+associated alignments and views are to be used to create the set of
+correspondences. This menu is useful when composing complex superpositions
+involving multi-domain and multi-chain complexes, when correspondences may be
+defined by more than one alignment.
+
+Note that these menu options appear when you have two or more structures in one Jmol viewer.
+
+\begin{figure}[htbp]
+\begin{center}
+\includegraphics[width=5.5in]{images/fdxsuperposition.pdf}
+\caption{{\bf Superposition of two ferredoxin structures.} The alignment on the
+left was used by Jalview to superpose structures associated with the
+FER1\_SPIOL and FER1\_MAIZE sequences in the alignment. Parts of each structure
+used for superposition are rendered as a cartoon, the remainder rendered in
+wireframe. The RMSD between corresponding positions in the structures before and
+after the superposition is shown in the Jmol console.}
+\label{mstrucsuperposition}
+\end{center}
+\end{figure}
+
+\subsection{Colouring Structure Data Associated with Multiple Alignments and
+Views} Normally, the original view from which a particular structure view was
+opened will be the one used to colour structure data. If alignments involving
+sequences associated with structure data shown in a Jmol have multiple views, Jalview gives you full control
+over which alignment, or alignment view, is used to colour the structure
+display. Sequence-structure colouring associations are
+changed {\sl via} the {\sl View $\Rightarrow$ Colour by ..} menu, which lists all
+views associated with data shown in the embedded Jmol view. A tick is shown beside
+views currently used as colouring source, and moving the
+mouse over each view will bring it to the front of the alignment display,
+allowing you to browse available colour sources prior to selecting one. If the
+{\sl Select many views} option is selected, then multiple views can be selected as sources for colouring the
+structure data. {\sl Invert selection} and {\sl Select all views} options are also provided to quickly change between multi-view selections.
+
+Note that the {\sl Select many views} option is useful if you have different
+views that colour different areas or domains of the alignment. This option is
+further explored in Section \ref{complexstructurecolours}.
+
+\begin{figure}[htbp]
+\begin{center}
+\includegraphics[width=5.5in]{images/mviewstructurecol.pdf}
+\caption{{\bf Choosing a different view for colouring a structure display}
+Browsing the {\sl View $\Rightarrow$ Colour by ..} menu provides full control
+of which alignment view is used to colour structures when the {\sl Colours
+$\Rightarrow$ By Sequence} option is selected.}
+\label{mviewstructurecol}
+\end{center}
+\end{figure}
+
+\exercise{Aligning Structures using the Ferredoxin
+Sequence Alignment}{\label{superpositionex}
+
+\exstep{Continue with the Jalview project created in exercise
+\ref{viewingstructex}. Use the {\sl Discover PDB IDs} function to retrieve PDB
+IDs associated with the FER1\_MAIZE sequence.}
+\exstep{Once discovery has completed, use the {\sl
+View PDB Structure} submenu to view one of the PDB file associated with
+FER1\_MAIZE (eg. 3B2F), Jalview will give you the option of aligning the
+structure to the one already open. To superimpose the structure associated with FER1\_MAIZE with the one
+associated with FER1\_SPIOL, press the {\sl Yes} button.
+
+{\sl The Jmol view will update to show both structures, and one will be
+moved on to the other. If this doesn't happen, use the Align function in the Jmol submenu}}
+\exstep{Create a new view on the alignment, and hide all but columns 121
+through to 132.}
+\exstep{Use the {\sl Jmol} submenu to
+recompute the superposition using just columns 121-132 of the alignment
+(The easiest way to achieve this is to select column 121-132 and in the View
+menu selected ``All but selected region'' from the Hide options).
+
+{\sl Note how the molecules shift position when superposed using a short part of
+the two structures.}}
+\exstep{Compare the initial and final RMSDs for superimposing molecules with
+the small section and with the whole alignment.}
+\exstep{The RMSD report can be
+viewed by right clicking the mouse on Jmol window, and select ``Show" and
+``Measurements".
+
+Which view do you think give the best 3D
+superposition, and why ?} }
+
+\subsubsection{Colouring Complexes}
+\label{complexstructurecolours}
+The ability to control which multiple alignment view is used to colour
+structural data is essential when working with data relating to
+multidomain biomolecules and complexes.
+
+In these situations, each chain identified in the structure may have a different
+evolutionary history, and a complete picture of functional variation can
+only be gained by integrating data from different alignments on the same
+structure view. An example of this is shown in Figure
+\ref{mviewalcomplex}, based on data from Song et. al.\footnote{Structure of
+DNMT1-DNA Complex Reveals a Role for Autoinhibition in Maintenance DNA Methylation. Jikui Song, Olga Rechkoblit, Timothy H. Bestor, and Dinshaw J. Patel.
+{\sl Science} 2011 {\bf 331} 1036-1040
+\href{http://www.sciencemag.org/content/331/6020/1036}{DOI:10.1126/science.1195380}}
+
+\begin{figure}[htbp]
+\begin{center}
+\includegraphics[]{images/mchainstructureview.pdf}
+\caption{{\bf The biological assembly of Mouse DNA Methyltransferase-1 coloured
+by Pfam alignments for its major domains} Alignments for each domain within the
+Uniprot sequence DNMT1\_MOUSE have been used to visualise sequence conservation
+in each component of this protein-DNA complex. Instructions for recreating this figure are given in exercise \ref{dnmtcomplexex}. }
+\label{mviewalcomplex}
+\end{center}
+\end{figure}
+
+\exercise{Colouring a Protein Complex to Explore Domain-Domain
+Interfaces}{\label{dnmtcomplexex}
+
+\exstep{Download the PDB file at
+\textsf{\url{http://www.jalview.org/tutorial/DNMT1\_MOUSE.pdb}} to your desktop.
+
+This is the biological unit for PDB ID 3pt6, as identified by the PDBe's PISA
+server.}
+\exstep{Launch the Jalview desktop and ensure you have at least 256MB of
+free memory available.
+{\sl Use the following webstart link:
+\href{http://www.jalview.org/services/launchApp?jvm-max-heap=1G}{http://www.jalview.org/services/launchApp?jvm-max-heap=1G}.}
+{\sl Alternatively in the Development section of the Jalview web site
+(\href{http://www.jalview.org/development/development-builds}{http://www.jalview.org/development/development-builds})
+in the ``latest official build'' row in the table, go to the
+``Webstart'' column, click on ``G2''.}}
+\exstep{Retrieve the following {\bf full} PFAM alignments: PF02008, PF01426
+(make sure you select the {\sl PFAM {\bf (Full)}} source). These will each be retrieved into their own alignment window.} \exstep{Drag the URL or file of the structure you downloaded in
+step 1 onto one of the alignments to associate it with the mouse sequence in
+that Pfam domain family.}
+\exstep{For every DNMT1\_MOUSE sequence in the alignment, use the sequence
+ID popup menu's {\sl Structure} submenu to view the DNMT1\_MOUSE structure for the associated mouse sequence. When given the option, {\bf view all of the structures in the same Jmol viewer}. Check the contents of the {\sl View $\Rightarrow$ Colour by ..} submenu to see what alignments can be used to
+colour the sequence.}
+\exstep{Repeat the previous two steps for each of
+the other alignments. In each case, when performing the `View DNMT1\_MOUSE.pdb'
+step, Jalview will ask if you wish to create a new Jmol view. You should
+respond `No', {\bf ensuring that each sequence fragment is associated with the same Jmol view}.}
+\exstep{Pick a different colourscheme for each alignment, and use the {\sl
+Colour by ..} submenu to ensure they are all used to colour the complex shown
+in the Jmol window.}
+\exstep{The final step needed to reproduce the shading in Figure
+\ref{mviewalcomplex} is to use the {\sl Colour $\Rightarrow$ By
+Annotation } option in each alignment window to shade the alignment by the
+{\bf Conservation} annotation row. This function was described in section
+\ref{colourbyannotation}.
+
+Ensure that you first disable the {\sl View $\Rightarrow$ Show Features} menu option, or you may not see any colour changes in the associated structure.