Mercurial > repos > iuc > jbrowse
changeset 29:5614cf75bb24 draft
Uploaded
| author | iuc |
|---|---|
| date | Thu, 31 Dec 2015 11:51:40 -0500 |
| parents | 8b1e63da255e |
| children | 12d5699f92e9 |
| files | gff3_rebase.py jbrowse-fromdir.xml jbrowse.xml macros.xml static/images/bam.png static/images/bigwig.png static/images/blast.png static/images/opacity.png static/images/sections.png static/images/styling.png tool_dependencies.xml |
| diffstat | 11 files changed, 303 insertions(+), 52 deletions(-) [+] |
line wrap: on
line diff
--- a/gff3_rebase.py Wed Dec 30 18:09:50 2015 -0500 +++ b/gff3_rebase.py Thu Dec 31 11:51:40 2015 -0500 @@ -145,7 +145,6 @@ subfeatures=False): new_subfeatures = child_features[feature.id] - # TODO: update starts fixed_subfeatures = [] for x in new_subfeatures: # Then update the location of the actual feature
--- a/jbrowse-fromdir.xml Wed Dec 30 18:09:50 2015 -0500 +++ b/jbrowse-fromdir.xml Thu Dec 31 11:51:40 2015 -0500 @@ -1,4 +1,4 @@ -<tool id="jbrowse-upgrade" name="JBrowse - Data Directory to Standalone" version="0.3"> +<tool id="jbrowse_to_standalone" name="JBrowse - Data Directory to Standalone" version="0.4"> <description>upgrades the bare data directory to a full JBrowse instance</description> <macros> <import>macros.xml</import> @@ -14,7 +14,7 @@ ## Copy in data directory cp -R $input.dataset.extra_files_path/data/ $output.files_path/ && -cat $output.files_path/index.html > $output +mv $output.files_path/index.html $output ]]></command> <inputs> <param label="JBrowse Data Directory"
--- a/jbrowse.xml Wed Dec 30 18:09:50 2015 -0500 +++ b/jbrowse.xml Thu Dec 31 11:51:40 2015 -0500 @@ -1,4 +1,4 @@ -<tool id="jbrowse" name="JBrowse" version="0.3"> +<tool id="jbrowse" name="JBrowse" version="0.4"> <description>genome browser</description> <macros> <import>macros.xml</import> @@ -23,15 +23,15 @@ --outdir $output.files_path; #if str($standalone) == "Complete": - cat $output.files_path/index.html > $output; + mv $output.files_path/index.html $output; #else: - cat $dummyIndex > $output; + mv $dummyIndex $output; #end if ## Ugly testing hack since I cannot get <extra_files> to test the files I want to test. Hmph. #if str($uglyTestingHack) == "enabled": - cp $trackxml $output + mv $trackxml $output #end if ]]></command> <configfiles> @@ -339,7 +339,259 @@ <data format="html" name="output" label="JBrowse on $on_string - $standalone"/> </outputs> <help><![CDATA[ -Build a static JBrowse visualization of a genome and some associated datasets. +JBrowse-in-Galaxy +================= + +JBrowse-in-Galaxy offers a highly configurable, workflow-compatible +alternative to Trackster. + +Overview +-------- + +JBrowse is a fast, embeddable genome browser built completely with +JavaScript and HTML5. + +The JBrowse-in-Galaxy (JiG) tool was written to help build complex +JBrowse installations straight from Galaxy, taking advantage of the +latest Galaxy features such as dataset collections, sections, and colour +pickers. It allows you to build up a JBrowse instance without worrying +about how to run the command line tools to format your data, and which +options need to be supplied and where. Additionally it comes with many +javascript functions to handle colouring of features which would be +nearly impossible to write without the assistance of this tool. + +The JBrowse-in-Galaxy tool is maintained by `Eric +Rasche <mailto:esr+jig@tamu.edu>`__, who you can contact if you +encounter missing features or bugs. + +Options +------- + +The first option you encounter is the **Fasta Sequence(s)**. This option +now accepts multiple fasta files, allowing you to build JBrowse +instances that contain data for multiple genomes or chrosomomes +(generally known as "landmark features" in gff3 terminology.) Up to 30 +will be shown from the dropdown selector within JBrowse, this is a known +issue. + +**Standalone Instances** are a somewhat in-development feature. +Currently Galaxy copies the entire JBrowse directory in order to have a +complete, downloadable file that contains a ready-to-go JBrowse +instance. This is obviously an anti-feature because users don't want a +complete copy of JBrowse (6-20Mb) that's duplicated for every JBrowse +dataset in their history, and admins don't want useless copies of +JBrowse on disk. Unfortunately we have not come up with the perfect +solution just yet, but we're working on it! In the meantime, users have +been given the option to produce just the ``data/`` directory. For those +unfamiliar with JBrowse, the ``data/`` directory contains processed data +files, but no way to view them. This feature is additionally implemented +for upcoming `Apollo <https://github.com/gmod/apollo>`__ integration. + +**Genetic Code** is a new feature in v0.4 of JiG / v1.12.0 of JBrowse, +which allows users to specify a non standard genetic code, and have +JBrowse highlight the correct start and stop codons. If you would like +to use a coding table not provided by this list, please let +`me <mailto:esr+jig@tamu.edu>`__ know so that I may add support for +this. + +**Track Groups** represent a set of tracks in a single category. These +can be used to let your users understand relationships between large +groups of tracks. + +.. image:: sections.png + +Annotation Tracks +----------------- + +Within Track Groups, you have one or more **Annotation Tracks**. Each +Annotation Track is a groups of datasets which have similar styling. +This allows you to rapidly build up JBrowse instances without having to +configure tracks individually. A massive improvement over previous +versions. For example, if you have five different GFF3 files from +various gene callers that you wish to display, you can take advantage of +this feature to style all of them similarly. + +There are a few different types of tracks supported, each with their own +set of options: + +GFF3/BED/GBK +~~~~~~~~~~~~ + +These are your standard feature tracks. They usually highlight genes, +mRNAs and other features of interest along a genomic region. The +underlying tool and this help documentation focus primarily on GFF3 +data, and have not been tested extensively with other formats. Automatic +min/max detection will likely fail under BED and GBK datasets. + +The data may be of a subclass we call **match/match part** data. This +consists of top level ``match`` features, with a child ``match_part`` +feature, and is often used in displaying alignments. (See "Alignments" +section on the `GFF3 +specification <http://www.sequenceontology.org/gff3.shtml>`__ for more +information). If the data is match/match part, you will need to specify +the top level match feature name, as it can be one of a few different SO +terms, and JiG does not yet have the ability to understand SO terms. + +Next up is the **Styling Options** section, which lets you control a few +properties on how the track is styled. Most of these you will not need +to configure and can safely leave on defaults. Occasionally you will +want to change what information is shown in the end product. + +.. image:: styling.png + +In the above image you can see some black text, and some blue text. The +source of the black text is configured with the **style.label** option, +and the source of the blue text is configured with the +**style.description** option. + +Feature Score Scaling & Colouring Options +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +First, you need to choose between ignoring the score attribute of GFF3 +files, or using it. If you choose to ignore it, all features will be +coloured with a solid colour. If you choose to use it, features will +have slightly different colours based on their scores. + +.. image:: opacity.png + +If you choose **Ignore score**, you may choose between automatically +choosing a colour, or manually specifying one. The automatically chosen +colours vary along a brewer palette and generally look quite nice with +no human intervention required. The manual colour choice is somewhat +self explanatory. Clicking on the small coloured square will bring up a +colour palette. + +If you choose **Base on score**, you're faced with a dizzying array of +options. First is the function to map the colour choices to colour +values. JiG comes with a few functions built in such as linear scaling, +logarithmic scaling, and blast scaling. + +The **linear scaling** method says "take these values, and they map +directly to a range of output values". **Logarithmic scaling** says +"please take the log of the score before mapping", and **Blast scaling** +is further specialised to handle blast data more nicely. These are +convenience functions to help transform the wide array of possible +values in the GFF3 score attribute to more meaningful numbers. If you +need more comprehensive score scaling, it is recommended that you +pre-process your GFF3 files somehow. + +Once you've selected a scaling method, you can choose to manually +specify the minimum and maximum expected values, or you can let JiG +determine them for you automatically. + +Finally, opacity is the only mapping we currently provide. Future +iterations will attempt to improve upon this and provide more colour +scales. The Opacity option maps the highest scoring features to full +opacity, and everything else to lower ones. + +BAM Pileups +~~~~~~~~~~~ + +We support BAM files and can automatically generate SNP tracks based on +that bam data. + +.. image:: bam.png + +This is *strongly discouraged* for high coverage density datasets. +Unfortunately there are no other configuration options exposed for bam +files. If you find JBrowse options you wish to see exposed, please let +`me <mailto:esr+jig@tamu.edu>`__ know. + +BlastXML +~~~~~~~~ + +.. image:: blast.png + +JiG now supports both blastn and blastp datasets. JiG internally uses a +blastXML to gapped GFF3 tool to convert your blastxml datasets into a +format ammenable to visualization in JBrowse. This tool is also +available separately from the IUC on the toolshed. + +**Minimum Gap Size** reflects how long a gap must be before it becomes a +real gap in the processed gff3 file. In the picture above, various sizes +of gaps can be seen. If the minimum gap size was set much higher, say +100nt, many of the smaller gaps would disappear, and the features on +both sides would be merged into one, longer feature. This setting is +inversely proportional to runtime and output file size. *Do not set this +to a low value for large datasets*. By setting this number lower, you +will have extremely large outputs and extremely long runtimes. The +default was configured based off of the author's experience, but the +author only works on small viruses. It is *strongly* recommended that +you filter your blast results before display, e.g. picking out the top +10 hits or so. + +**Protein blast search** option merely informs underlying tools that +they should adjust feature locations by 3x. + +Styling Options +^^^^^^^^^^^^^^^ + +Please see the styling options for GFF3 datasets, they are identical. + +Feature Score Scaling & Coloring Options +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +Please see the score scaling and colouring options for GFF3 datasets, +they are identical. Remember to set your score scaling to "blast" method +if you do use it. + +Bigwig XY +~~~~~~~~~ + +.. image:: bigwig.png + +**XYPlot** + +BigWig tracks can be displayed as a "density" plot which is continuous +line which varies in colour, or as an "XYplot." XYplots are preferrable +for users to visually identify specific features in a bigwig track, +however density tracks are more visually compact. + +**Variance Band** is an option available to XYPlots, and can be seen in +the third and fourth tracks in the above picture. This overlays a mean +line, and 1 and 2 standard deviation areas. + +**Track Scaling** is different from colour scaling, instead it +configures how the track behaves inside of JBrowse. **Autoscaling +globally** means that JBrowse will determine the minimum and maximum for +the track, and fix the bounds of the viewport to that. E.g. if your +track ranges from 1-1000, and the region you're currently zoomed to only +goes from 0-50, then the viewport range will still show 1-1000. This is +good for global genomic context. However you may wish to consider +**autoscaling locally** instead. In the example of a region which varies +from 0-50, autoscaling locally would cause the individual track's +viewport to re-adjust and show just the 0-50 region. If neither of these +options are paletable, you may manually hardcode the minimum and +maximums for the track to scale to. + +Colour Options +^^^^^^^^^^^^^^ + +BigWig tracks have two colours in JBrowse, a positive and a negative +colour. + +As always you may manually choose a colour, or let JiG choose for you. + +One of the more interesting options is the **Bicolor pivot**. This +option allows you to control the point at which JBrowse switches from +the positive colour to the negative. In the above graphic, you can see +this has been configured to "mean" for the first two (orange and blue) +tracks. + +VCFs/SNPs +~~~~~~~~~ + +These tracks do not support any special configuration. + +Known Issues +------------ + +- More than 30 landmark features cannot be listed in the manual + selector. +- Non GFF3 likely has issue with automatically determined min/max + scores. Manually specify minimum and maximum score attributes, or do + not use varied colours based on scores to avoid this issue. + @ATTRIBUTION@ ]]></help>
--- a/macros.xml Wed Dec 30 18:09:50 2015 -0500 +++ b/macros.xml Thu Dec 31 11:51:40 2015 -0500 @@ -25,7 +25,7 @@ <token name="@ATTRIBUTION@"><![CDATA[ **Attribution** -This Galaxy tool relies on the JBrowse, maintained by the GMOD Community +This Galaxy tool relies on the JBrowse, maintained by the GMOD Community. The Galaxy wrapper is developed by Eric Rasche ]]> </token> <xml name="auto_manual_tk" @@ -64,19 +64,19 @@ </xml> <xml name="auto_color" - token_cond_label="Color" - token_cond_name="color" - token_select_label="Color Selection" - token_select_name="color_select" - token_automatic_label="Automatically selected" - token_manual_label="Manual Color Selection"> + token_cond_label="Color" + token_cond_name="color" + token_select_label="Color Selection" + token_select_name="color_select" + token_automatic_label="Automatically selected" + token_manual_label="Manual Color Selection"> <expand macro="auto_manual_tk" - cond_label="@COND_LABEL@" - cond_name="@COND_NAME@" - select_label="@SELECT_LABEL@" - select_name="@SELECT_NAME@" - automatic_label="@AUTOMATIC_LABEL@" - manual_label="@MANUAL_LABEL@"> + cond_label="@COND_LABEL@" + cond_name="@COND_NAME@" + select_label="@SELECT_LABEL@" + select_name="@SELECT_NAME@" + automatic_label="@AUTOMATIC_LABEL@" + manual_label="@MANUAL_LABEL@"> <expand macro="jb_color" /> <yield /> </expand> @@ -110,15 +110,15 @@ token_automatic_label="Automatically selected" token_manual_label="Manual Color Selection"> <expand macro="jb_color" - label="JBrowse style.pos_color" - name="style_pos_color" - value="blue" - help="CSS color, default 'blue'. When drawing bicolor plots, the fill color to use for values that are above the pivot point." /> + label="JBrowse style.pos_color" + name="style_pos_color" + value="blue" + help="CSS color, default 'blue'. When drawing bicolor plots, the fill color to use for values that are above the pivot point." /> <expand macro="jb_color" - label="JBrowse style.neg_color" - name="style_neg_color" - value="red" - help=" CSS color, default 'red'. When drawing bicolor plots, the fill color to use for values that are below the pivot point." /> + label="JBrowse style.neg_color" + name="style_neg_color" + value="red" + help=" CSS color, default 'red'. When drawing bicolor plots, the fill color to use for values that are below the pivot point." /> </expand> <conditional name="bicolor_pivot" label="Bicolor Pivot"> @@ -137,8 +137,8 @@ </xml> <xml name="color_selection" - token_scaling_lin_select="true" - token_scaling_log_select="false" + token_scaling_lin_select="true" + token_scaling_log_select="false" > <section name="jbcolor_scale" title="JBrowse Feature Score Scaling & Coloring Options [Advanced]" expanded="false"> <conditional name="color_score" label="JBrowse style.color & Score relationship"> @@ -154,7 +154,7 @@ <when value="score"> <!-- Scaling --> <param type="select" label="JBrowse style.color function's score scaling" name="score_scaling" - help="How should the colors be distributed across the values? For blast results which distributes scores on the scale of approximately [1e-500, 10], it makes sense to request a logarithmic scaling of the color values. Logarithmic is indeed the default for blast. However other analysis methods may produce scores on ranges such as [0, 100] where a linear scale would be more appropriate for color distribution."> + help="How should the colors be distributed across the values? For blast results which distributes scores on the scale of approximately [1e-500, 10], it makes sense to request a logarithmic scaling of the color values. Logarithmic is indeed the default for blast. However other analysis methods may produce scores on ranges such as [0, 100] where a linear scale would be more appropriate for color distribution."> <option value="linear" selected="@SCALING_LIN_SELECT@" >Linear scaling</option> <option value="logarithmic" selected="" >Logarithmic scaling</option> <option value="blast" selected="@SCALING_LOG_SELECT@" >Blast scaling</option> @@ -194,30 +194,30 @@ </section> </xml> <xml name="track_styling" - token_classname="feature" - token_label="name,id" - token_description="note,description" - token_height="100px"> + token_classname="feature" + token_label="name,id" + token_description="note,description" + token_height="100px"> <section name="jbstyle" title="JBrowse Styling Options [Advanced]" expanded="false"> <param label="JBrowse style.className" type="text" - name="style_classname" - value="@CLASSNAME@" - help="Set CSS style.className property"/> + name="style_classname" + value="@CLASSNAME@" + help="Set CSS style.className property"/> <param label="JBrowse style.label" - type="text" - name="style_label" - value="@LABEL@" - help="Comma-separated list of case-insensitive feature tags to use for showing the feature's label. The first one found will be used. Default 'name,id'" /> + type="text" + name="style_label" + value="@LABEL@" + help="Comma-separated list of case-insensitive feature tags to use for showing the feature's label. The first one found will be used. Default 'name,id'" /> <param label="JBrowse style.description" - type="text" - name="style_description" - value="@DESCRIPTION@" - help="Comma-separated list of case-insensitive feature tags to check for the feature's long description. The first one found will be used."/> + type="text" + name="style_description" + value="@DESCRIPTION@" + help="Comma-separated list of case-insensitive feature tags to check for the feature's long description. The first one found will be used."/> <param label="JBrowse style.height" - type="text" - name="style_height" - value="@HEIGHT@" - help="Height in pixels of glyphs. Default value varies from glyph to glyph. Note that the 'compact' displayMode uses style->height * 0.35 so changing style height can adjust the compact visualization."/> + type="text" + name="style_height" + value="@HEIGHT@" + help="Height in pixels of glyphs. Default value varies from glyph to glyph. Note that the 'compact' displayMode uses style->height * 0.35 so changing style height can adjust the compact visualization."/> </section> </xml>
--- a/tool_dependencies.xml Wed Dec 30 18:09:50 2015 -0500 +++ b/tool_dependencies.xml Thu Dec 31 11:51:40 2015 -0500 @@ -1,7 +1,7 @@ <?xml version="1.0"?> <tool_dependency> <package name="jbrowse" version="1.12.0"> - <repository changeset_revision="0fdf1eabbc7d" name="package_jbrowse_1_12_0" owner="iuc" toolshed="https://testtoolshed.g2.bx.psu.edu" /> + <repository changeset_revision="469ecaf54ab6" name="package_jbrowse_1_12_0" owner="iuc" toolshed="https://testtoolshed.g2.bx.psu.edu" /> </package> <package name="bundle_jbrowse" version="1.12.0"> <repository changeset_revision="eafed51757f5" name="package_perl_bundle_jbrowse_1_12_0" owner="iuc" toolshed="https://testtoolshed.g2.bx.psu.edu" />