--- /dev/null
+<html>
+<head>
+<title>Modeller PIR Format IO</title>
+</head>
+<body>
+<p><strong>Modeller PIR Format IO</strong></p>
+<p>The homology modelling program, <a
+href="http://salilab.org/modeller/">Modeller</a> uses a special form
+of the PIR format where information about sequence numbering and chain
+codes are written into the 'description' line between the PIR protein
+tag and the protein alignment entry:</p>
+<pre>>P1;Q93Z60_ARATH
+sequence:Q93Z60_ARATH:1:.:118:.:.
+----MASTALSSAIVSTSFLRRQQTPISLRSLPFANT-QSLFGLKS-STARGGRVTAMATYKVKFITPEGEQ
+EVECEEDVYVLDAAEEAGLDLPYSCRAGSCSSCAGKVVSGSIDQSD------QSFLD-D-------------
+---------------------*
+>P1;PDB|1FER|_
+structureX:1FER:1:.:105:.:.
+----------------------------------------------------AFVVTDNCIKCKY---TDCV
+EV-CPVDCFY----EGPNFLVIHPDECIDCALCEPECPAQAIFSEDEVPEDMQEFIQLNAELAEVWPNITEK
+KDPLPDAEDWDGVKGKLQHLE*
+</pre>
+<p>Jalview will attempt to parse any PIR entries conforming to the
+Modeller/PIR format, in order to extract the sequence start and end
+numbering and (possibly) a PDB file reference. The description line
+information is always stored in the sequence description string - so
+no information is lost if this parsing process fails.</p>
+<p>The 'Modeller Output' flag in the 'Output' tab of the Jalview <a href="preferences.html">Preferences dialog
+box</a> controls whether Jalview will also output MODELLER style PIR
+files. In this case, any existing 'non-modeller PIR' header
+information in the description string of an alignment is appended to
+an automatically generated modeller description line for that
+sequence.</p>
+<p>The general format used for generating the Modeller/PIR sequence
+description line is shown below :
+<pre>>P1;<em>Primary_Sequence_ID</em>
+<em>sequence or structureX</em>:<em>pdb-reference if
+ available</em>:<em>start residue</em>:<em>start chain code</em>:<em>end
+ residue</em>:<em>end chain code</em>:. <em>description text</em></pre>
+The first field is either sequence or structureX, depending upon the
+presence of a PDB database ID for the sequence. If the protein has no
+PDB reference, then the chain code is not specified, unless one
+already existed when the sequence was imported into Jalview.
+</body>
+</html>