<p>
<strong>Principal Component Analysis</strong>
</p>
+ <p>
+ A principal component analysis can be performed via the <a
+ href="calculations.html">calculations dialog</a> which is accessed
+ by selecting <strong>Calculate→Calculate Tree or
+ PCA...</strong>.
+ </p>
<p>This calculation creates a spatial representation of the
similarities within a selected group, or all of the sequences in an
alignment. After the calculation finishes, a 3D viewer displays the
set of sequences as points in 'similarity space', and similar
sequences tend to lie near each other in the space.</p>
<p>
- <em>Caveats</em><br />The calculation is computationally expensive,
- and may fail for very large sets of sequences - usually because the
- JVM has run out of memory. A future release of Jalview will be able
- to avoid this by executing the calculation via a web service.
+ <em>Caveats</em><br />The calculation can be computationally
+ expensive, and may fail for very large sets of sequences - usually
+ because the JVM has run out of memory. However, the PCA
+ implementation in Jalview 2.10.2 employs more memory efficient
+ matrix storage structures, allowing larger PCAs to be performed.
</p>
<p>
calculation are given in the <strong><em>Change
Parameters</em></strong> menu.
</p>
- <p>
- <em>PCA Calculation modes</em><br /> The default Jalview
- calculation mode (indicated when <em><strong>Jalview
- PCA Calculation</strong></em> is ticked in the <strong><em>Change
- Parameters</em></strong> menu) is to perform a PCA on a matrix where elements
- in the upper diagonal give the sum of scores for mutating in one
- direction, and the lower diagonal is the sum of scores for mutating
- in the other. For protein substitution models like BLOSUM62, this
- gives an asymmetric matrix, and a different PCA to a matrix produced
- with the method described in the paper by G. Casari, C. Sander and
- A. Valencia. Structural Biology volume 2, no. 2, February 1995 (<a
- href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=7749921">pubmed</a>)
- and implemented at the SeqSpace server at the EBI. This method
- preconditions the matrix by multiplying it with its transpose, and
- can be employed in the PCA viewer by unchecking the <strong><em>Jalview
- PCA Calculation</em></strong> option in the <strong><em>Change
- Parameters</em></strong> menu.
- </p>
<img src="pcaviewer.gif">
<p>
<strong>The PCA Viewer</strong>
added to the Jalview desktop in v2.7.</em> <em>The Reset button
and Change Parameters menu were added in Jalview 2.8.</em> <em>Support
for PAM250 based PCA was added in Jalview 2.8.1.</em>
+ </p>
+ <p>
+ <strong>Reproducing PCA calculations performed with older
+ Jalview releases</strong> Jalview 2.10.2 included a revised PCA
+ implementation which treated Gaps and non-standard residues in the
+ same way as a matrix produced with the method described in the paper
+ by G. Casari, C. Sander and A. Valencia. Structural Biology volume
+ 2, no. 2, February 1995 (<a
+ href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=7749921">pubmed</a>)
+ and implemented at the SeqSpace server at the EBI. To reproduce
+ calculations performed with earlier Jalview releases it is necessary
+ to execute the following Groovy script:
+ <pre>
+ jalview.analysis.scoremodels.ScoreMatrix.scoreGapAsAny=true
+ jalview.analysis.scoremodels.ScoreModels.instance.BLOSUM62.@matrix[4][1]=3
+ </pre>
+ This script enables the legacy PCA mode where gaps were treated as
+ 'X', and to modify the BLOSUM62 matrix so it is asymmetric for
+ mutations between C to R (this was a typo in the original Jalview
+ BLOSUM62 matrix which was fixed in 2.10.2).
+ </p>
</body>
</html>
git://source.jalview.org / jalview.git / blobdiff