3 * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
4 * Copyright (C) $$Year-Rel$$ The Jalview Authors
6 * This file is part of Jalview.
8 * Jalview is free software: you can redistribute it and/or
9 * modify it under the terms of the GNU General Public License
10 * as published by the Free Software Foundation, either version 3
11 * of the License, or (at your option) any later version.
13 * Jalview is distributed in the hope that it will be useful, but
14 * WITHOUT ANY WARRANTY; without even the implied warranty
15 * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
16 * PURPOSE. See the GNU General Public License for more details.
18 * You should have received a copy of the GNU General Public License
19 * along with Jalview. If not, see <http://www.gnu.org/licenses/>.
20 * The Jalview Authors are detailed in the 'AUTHORS' file.
23 <title>Working with PAE Matrices in Jalview</title>
28 <strong>Working with Predicted Alignment Error Matrices in
32 <p>Predicted Alignment Error (PAE) matrices are produced by
33 deep-learning based 3D-structure prediction pipelines such as
34 AlphaFold. They reflect how reliably two parts of a model have been
35 positioned in space, by giving for each residue the likely error (in
36 Ångstroms) between that residue and every other modelled
37 position the pair of residues' real relative position, if the model
38 and real 3D structure were superimposed at that residue.</p>
40 Jalview visualises PAE matrices as an alignment annotation track,
41 shaded from dark green to white, similar to the encoding used on the
42 EBI-AlphaFold website (see <a
43 href="https://alphafold.ebi.ac.uk/entry/O04090">O04090 3D model</a>
46 <div style="display: flex; flex-wrap: wrap;" align="center"
49 <img src="../structures/epas1_annotdetail.png" height="300" />
51 Alignment of EPAS1 homologs from Human, Rat and Cow<br />with
52 predicted alignment error shown for Human
56 <img src="../structures/epas1_pae_ebiaf.png" height="300" />
58 Predicted Alignment Error for Human EPAS1<br />from <a
59 href="https://alphafold.ebi.ac.uk/entry/Q99814">https://alphafold.ebi.ac.uk/entry/Q99814</a>
64 <strong>Importing PAE Matrices</strong>
67 Jalview retrieves PAE matrices when importing predicted 3D structures
68 from the EBI-AlphaFold database via <a
69 href="../features/structurechooser.html">Jalview's structure
70 chooser</a> GUI. If you have produced your own models and accompanying
71 PAE matrices using a pipeline such as ColabFold, then you can load
72 them both together via the <a
73 href="../features/structurechooser.html#loadpdbfile">Load PDB
74 File</a> dropdown menu in the 3D structure chooser, providing it is in a
75 <a href="../io/paematrixformat.html">supported PAE format</a>.
78 The <a href="../features/clarguments-basic.html">Command Line
79 Interface</a> also provides a options for importing PAE matrices along
80 side models, enabling the automated production of alignment figures
81 annotated with PAE matrices and PLDDT scores.
84 <strong>Showing PAE Matrix Annotations </strong>
87 When viewing 3D structures from the EBI-AlphaFold database or local 3D
88 structures with an associated PAE file, the PAE is imported as <i>Reference
89 Annotation</i>, which is not always automatically added to the alignment
92 <p>To show the PAE, right click the sequence and locate the 'Add
93 Reference Annotation' entry in the Sequence ID submenu, or select all
94 sequences and locate the option in the Selection submenu. You can do
95 this in any alignment window (or view) where a sequence with
96 associated PAE data appears.</p>
98 <strong>Adjusting the height of PAE matrix annotations</strong>
101 PAE annotations behave in the same way as Jalview's line graph and
102 histogram tracks. Click+dragging up and down with the left (select)
103 mouse button held down will increase or decrease the height of the
104 annotation. You can also hold down <strong><em>SHIFT</em></strong>
105 whilst doing this to adjust the height of all PAE rows at once.
107 <p>PAE matrix annotation rows behave like any other sequence
108 associated annotation, with the following additional features:</p>
110 <li>The vertical axis of the PAE heatmap is mapped to positions
111 on the linked 3D structure.
113 <li>Mousing over the matrix shows a tooltip giving information
114 on the range of values under the mouse.<br />Positions in the
115 associated 3D structure are also highlighted in any linked views.
117 <li>Clicking on positions in the matrix selects columns of the
118 alignment corresponding to the row and column in the matrix.</li>
121 <li>Rectangular selections (created by Cmd (or Alt)+Click
122 dragging on the matrix) can be created to select multiple ranges of
123 columns at once.</li>
124 <li>Columns corresponding to adjacent regions with similarly low
125 levels of predicted alignment error can be selected by Ctrl+Clicking
126 on a region in the matrix.</li>
127 <li>Columns of an alignment showing a PAE matrix can be grouped
128 and selected by clustering the matrix.</li>
131 <strong><a name="clustering">Clustering PAE Matrices</a></strong>
133 <p>PAE matrices are useful for identifying regions of 3D structure
134 predictions that are likely to be positioned in space in the same or
135 similar way as shown in the predicted structure data. Regions of low
136 PAE often correlate with high alphafold reliability (PLDDT) scores,
137 but also complement them since they highlight well-folded regions such
138 as domains, and how well those regions have been predicted to be
139 positioned relative to eachother, which is important when evaluating
140 whether domain-domain interactions or other contacts can be trusted.</p>
141 <p>To make it more easy to identify regions of low PAE, Jalview can
142 cluster the PAE matrix, allowing columns of the matrix to be grouped
143 according to their similarity, using an Average Distance (UPGMA) tree
144 algorithm and the sum of differences between each column's PAE values.</p>
146 <strong><em>dist<sub>ij</sub></em> = ∥ <em><u>p</u><sub>i</sub>-<u>p</u><sub>j</sub></em>
150 To create a PAE matrix tree, right click on a PAE annotation's label
151 to open the annotation popup menu, and select <strong><em>Cluster
152 Matrix</em></strong>. Once the calculation has finished, a tree viewer will open,
153 and columns of the matrix are then partitioned into groups such that
154 the third left-most node from the root is placed in its own group.
155 Colours are randomly assigned to each group, and by default these will
156 also be overlaid on the matrix annotation row.
159 <li>The PAE matrix tree viewer behaves like other tree views in
160 Jalview, except selecting nodes or groups of nodes in the tree select
161 columns in the alignment rather than sequences, and clicking adjust
162 the matrix's partition.</li>
163 <li>Only one tree and clustering can be defined for a PAE matrix,
164 regardless of whether it is displayed in different views or
166 <li>Double clicking on a position in the PAE annotation where a
167 clustering has been defined will select both the row and column
168 clusters for the clicked position. This makes it easy to select
169 clusters corresponding to pairs of interacting regions.</li>
170 <li>Cluster colours for a PAE matrix can be used to colour
171 sequences or columns of the alignment via the <strong><em><a
172 href="../colourSchemes/annotationColouring.html">Colour by
173 Annotation.. dialog</a></em></strong>
174 (opened by right-clicking the annotation label
175 and selecting from the popup menu).
179 <strong>PAE matrices and Jalview Projects</strong>
181 <p>Any PAE matrices imported to Jalview are saved along side any
182 trees and clustering defined on them in Jalview Projects.</p>
184 <em>Support for visualision and analysis of predicted alignment
185 error matrices was added in Jalview 2.11.3. </em>