Description

MotifLab provides several data formats to allow users to output their data and analysis results in many different ways. Nevertheless, there may be times when the functionality offered by these formats is not enough, for example if a user has run several analyses and created multiple data objects and wants to combine information from all of these in a single compact document with follows a predefined layout.
The TemplateHTML format (and its companion Template) can be used in such cases to achieve greater control over the presentation of the output. The data format requires a "template" document which must be provided in the form of a Text Variable. This template can contain regular text and HTML-markup which can be interspersed with references to named data objects on the form: {dataobject}. When the template Text Variable is output in TemplateHTML-format, the references to data objects in the text will be replaced with the actual contents of these data objects. No error message or warning will be given if the data object named in the reference does not exist or if there is some other mistake in the reference. In such cases, MotifLab will simply leave the original reference as is in the output. Note that only "simple" data objects such as Numeric Variables, Text Variables, Collections and OutputData objects can be referenced in the template. However, other types of data objects (such as feature data, partitions or maps) can be included by first outputting these to intermediate OutputData objects which can then be referenced.

The contents of a referenced data object will usually be included directly in a default format, but some control over the presentation is provided with additional options which are described in the documentation for the Template format.

Example:
In the following scenario, a user has created a protocol script to find significant transcription factor motifs in promoters for sets of sequences that are either up- or downregulated at three different timepoints (1h, 2h or 3h). The sequence collections are named Up1, Up2, Up3, Down1, Down2 and Down3, and the collections of significant motifs are called Motifs_Up1, Motifs_Up2, Motifs_Up3, Motifs_Down1, Motifs_Down2 and Motifs_Down3. What the user wants now is to create a simple report that contains information about how many sequences there were in each such collection and list the names of the significant TFs for each collection. The following protocol first creates six OutputData objects containing names of significant TFs and then creates the final report based on a predefined template.

Motifs_Up1_out = output Motifs_Up1 in Motif_Properties format {Format="Clean Short name"}
Motifs_Up2_out = output Motifs_Up2 in Motif_Properties format {Format="Clean Short name"}
Motifs_Up3_out = output Motifs_Up3 in Motif_Properties format {Format="Clean Short name"}
Motifs_Down1_out = output Motifs_Down1 in Motif_Properties format {Format="Clean Short name"}
Motifs_Down2_out = output Motifs_Down2 in Motif_Properties format {Format="Clean Short name"}
Motifs_Down3_out = output Motifs_Down3 in Motif_Properties format {Format="Clean Short name"}

TemplateText = new Text Variable(File:"Template_text.html", format=Plain)
Output1= output TemplateText in TemplateHTML format

The file "Template_text.html" which is used in the protocol above to provide the template text for the final report is shown below. All references to data objects in the template are marked in red color. First the template produces a 2x3-table with the sizes of each of the sequence collections. Then it writes out a second table listing the significant transcription factors for each such collection. It would be possible to reference the motif collections directly in the template, e.g. {Motifs_Up1}. However, this would then only list the IDs of the motifs and not the TF-names, so instead the protocol above uses the Motif_Properties data format to output the "clean short name" of each motif in each collection to an intermediate OutputData object that is referenced instead (Motifs_Up1_out). These OutputData objects contains one TF-name on each line and there might be duplicate names if there are several motif models for the same TF. Hence, the template specifies that the names in the OutputData objects should be sorted alphabetically and duplicate names should be removed (using the "AU" option).

<h2>Number of genes up- and down-regulated at different time points:</h2>

<table>
    <tr><th>Time</th><th>1h</th><th>2h</th><th>3h</th></tr>
    <tr><td>Up</td><td>{Up1:size}</td><td>{Up2:size}</td><td>{Up3:size}</td></tr>
    <tr><td>Down</td><td>{Down1:size}</td><td>{Down2:size}</td><td>{Down3:size}</td></tr>
</table>

<br>

<h2>Significant transcription factors in promoters of these genes:</h2>

<table>
<tr>
    <th style="background-color:#E0E0E0;">1h</th>
    <th style="background-color:#C8C8C8;">2h</th>
    <th style="background-color:#B0B0B0;">3h</th>
</tr>
<tr>
    <td valign=top style="background-color:#FFD0D0;">{Motifs_Up1_out::AU}</td>
    <td valign=top style="background-color:#FFC0C0;">{Motifs_Up2_out::AU}</td>
    <td valign=top style="background-color:#FFC0B0;">{Motifs_Up3_out::AU}</td>
</tr>
<tr>
    <td valign=top style="background-color:#D0FFD0;">{Motifs_Down1_out::AU}</td>
    <td valign=top style="background-color:#C0FFC0;">{Motifs_Down2_out::AU}</td>
    <td valign=top style="background-color:#B0FFB0;">{Motifs_Down3_out::AU}</td>
</tr>
</table>

The result could look something like this:

Number of genes up- and down-regulated at different time points:

Time	1h	2h	3h
Up	23	42	31
Down	17	35	26

Significant transcription factors in promoters of these genes:

1h	2h	3h
EGR1 FOS1	FOXD1 GATA2 HOXA5 IRX5 JUND	ATF5 NFX1 NKX3-1 ZHX2
	E2F5 KLF15 NR2C1	EVI1 HEYL HOXC9 POU2F1