Skip to content

Latest commit

 

History

History
340 lines (213 loc) · 10.2 KB

findSRRs.md

File metadata and controls

340 lines (213 loc) · 10.2 KB
Raw

The tool: Find-SRRs

The Find-SRRs tool is a perl script which can be found here: https://raw.githubusercontent.com/statonlab/Finding-SSRs/master/findSSRs_altered.pl. This instruction shows how to wrap this perl script into a Aurora Galaxy Tool.

Launch GTG and Galaxy

Get a copy of launch_dev_env.sh to your local computer

wget https://raw.githubusercontent.com/MingChen0919/gtgdocker/master/launch_dev_env.sh

Run launch_dev_env.sh

sh launch_dev_env.sh

This will start the GTG app, a Galaxy instance, and creates a workspace with three folders within it.

gtg_dev_dir
├── database
├── galaxy_tool_repository
└── shed_tools

Create initial XML

We will wrap findSSRs_altered.pl as an Aurora Galaxy Tool.

  • Click Create Tool XML and Fill the XML file name field with rmarkdown_report.xml
  • Select Aurora Galaxy Tool and click Save

Edit the tool and tool->description components to give your tool a meaningful tool name and tool description.

Add components to XML

Add requirement components

This tool requires three perl modules - Getopt::Long, Bio::SeqIO, and Excel::Writer::XLSX - as well as Primer3. We create requirement component for each of these four components. We can find the dependencies and their version numbers in bioconda.

  • perl-getopt-long (2.50)
  • perl-bioperl (1.7.2)
  • perl-excel-writer-xlsx (0.98)
  • primer3 (2.4.1a)

Add input components

The Find SRRs only has two inputs. We create input component for each.

# Usage: findSSRs.pl <arguments>
#
# The list of arguments includes:
#
# -f|--fasta_file <fasta_file>
# Required.  The file of the sequences to be searched.
#
# -m|--masked_file <masked_fasta_file>
# Required.  A soft-masked version of the fasta file (soft masked means low
# complexity sequences are in lower case bases.)

The fasta_file input

Add a tool->inputs->param(type: data) component because this input takes a file path.

Edit input attributes

The masked_file input

Add a tool->inputs->param(type: data) component because this input takes a file path.

Edit input attributes

Add output components

Since we are building an Aurora Galaxy Tool, the tool report will provide access to all tool outputs. However, if we want to expose some outputs to the Galaxy history so that it can be used as inputs for other tools, we will need to create output components for those outputs.

Below is what this program outputs:

# Output:
# ------
# <input-file-name>.ssr.fasta
# A fasta file with sequences with a SSR. (Sequences with compound SSRs are included)
#
# <input-file-name>.ssr_stats.txt
# A text file of statistics about the SSRs discovered.
#
# <input-file-name>.ssr_report.txt
# A tab-delimited file with each SSR.  The columns are SSR ID,
# motif, number of repeats, start position, end position.
#
# <input-file-name>.ssr_report.xlsx
# A excel file with SSR results and stats
#
# <input-file-name>.di_primer_report.txt
# <input-file-name>.tri_primer_report.txt
# <input-file-name>.tetra_primer_report.txt
# Tab-delimited files with sequences with a specified SSR motif length.  Columns are
# SSR ID, motif, number of repeats, start position, end position, left primer,
# right primer, left primer Tm, right primer Tm, amplicon size

In this Galaxy tool, I choose to expose the following three outputs to the Galaxy history:

  • <input-file-name>.ssr.fasta
  • <input-file-name>.ssr_stats.txt
  • <input-file-name>.ssr_report.txt

The ssr_fasta output

Add a tool->outputs->data component.

Edit output attributes

The ssr_stats output

Add a tool->outputs->data component.

Edit output attributes

The ssr_report output

Add a tool->outputs->data component.

Edit output attributes

Edit XML command section

We edit the shell script within the command section. In Aurora Galaxy Tools, this shell script is only used to collect user inputs and any command line arguments.

Instead of editing the shell script directly through the component web form, we edit the shell script through the gtg_dev_dir/galaxy_tool_repository/rmarkdown_report.sh file. The content of this file will be automatically plugged into the command section when the XML page is being viewed.

Open gtg_dev_dir/galaxy_tool_repository/rmarkdown_report.sh and add the following content to the end of the file.

            -f $fasta_file
            -m $masked_file

            -F $ssr_fasta
            -S $ssr_stats
            -R $ssr_report

After completing the edit of rmarkdown_report.sh file, click the View/Update XML tab twice to integrate the updates.

Below is the updated command section

Edit getopt.csv

In Aurora Galaxy Tools, the rmarkdown_report_render.R script runs as a command line program. The main purpose of it is to pass user inputs to rmarkdown_report.Rmd and render rmarkdown_report.Rmd to execute data analysis and generate analysis report. The getopt.csv is used to specify the short and long flags/options accepted by rmarkdown_report_render.R.

Open getopt.csv and add the following content to the csv file.

short flag	argument mask	data type	variable name
o			1				character	report
d			1				character	report.files_path
f			1				character	fasta_file
m			1				character	masked_file
F			1				character	ssr_fasta
S			1				character	ssr_stats
R			1				character	ssr_report

The short flag and the variable name column should be the same as what you put in the command section. Details can be found here about how to specify options/flags: https://cran.r-project.org/web/packages/getopt/getopt.pdf.

How to use the variables collected by rmarkdown_report_render.R?

Aurora Galaxy Tools passes the arguments with two options: one is a set of R variables stored in a list named opt; the other is a set of environment variables. Variables are named in the format of X_SHORT_FLAG. For example, if I want to use the the variable $report.files_path in an R code chunk, I can access it by opt$X_d. If I want to use it as a environment variable, I can access it by $X_d.

Edit rmarkdown_report.Rmd

The rmarkdown_report.Rmd is where all the analysis logic should go. First, you need to download the findSSRs_altered.pl script and place it into the gtg_dev_dir/galaxy_tool_repository directory.

cd gtg_dev_dir/galaxy_tool_repository
wget https://raw.githubusercontent.com/statonlab/Finding-SSRs/master/findSSRs_altered.pl

Add the content below to the end of the rmarkdown_report.Rmd file.

# Run `findSRRs_altered.pl`

* `${REPORT_FILES_PATH}`: the output directory that stores all analysis outputs.
* `${TOOL_INSTALL_DIR}`: the tool install directory where the `findSSRs_altered.pl` is located.
* `$X_f`: the path to the fasta file.
* `$X_m`: the path to the masked fasta file.

```{bash}
cd ${REPORT_FILES_PATH}
perl ${TOOL_INSTALL_DIR}/findSSRs_altered.pl \
  -f $X_f \
  -m $X_m
```

Publish tool to Test ToolShed

Add Test ToolShed API key

GTG uses planemo in the backend to publish tools to ToolShed, which requires ToolShed API keys. Below is the interface to add ToolShed API keys.

Build Tool Repository

Once we complete building the XML file, we need to add it to the gtg_dev_dir/galaxy_tool_repository. Click Build Tool Repository, select rmarkdown_report.xml and then click Update XMLs in galaxy_tool_directory folder.

.

We also need to unzip the vakata-jstree-3.3.5.zip file within gtg_dev_dir/galaxy_tool_repository directory. After the file is unzipped, we can delete the vakata-jstree-3.3.5.zip file.

Publish tool

Below is the webform to publish the tool to ToolShed or Test ToolShed. Fill out the form with necessary tool meta data and then click Publish to Test Tool Shed to publish the tool to the Test ToolShed first.

Test Tool in Galaxy

Install Tool into Galaxy

Login to the connected Galaxy (http://127.0.0.1:8081/) with the username and password below:

  • user: admin@galaxy.org
  • password: admin

Click admin->Search Tool Shed->Galaxy Test Tool Shed and search for tool aurora_findssrs and install the tool.

Test aurora_findssrs

Sync tool in GTG to tool installed in Galaxy

Link tool between GTG and Galaxy

Developing Galaxy tools is an iterative process. It's very likely that you need to update your XML file frequently. GTG allows you to test the updates easily by syncing the tool in GTG into the connected Galaxy instance. You will use the following interface to link the tool in GTG with the same tool installed in Galaxy. The path to the tool can be found in the gtg_dev_dir/shed_tools. For example, in this case the path would be shed_tools/testtoolshed.g2.bx.psu.edu/repos/mingchen0919/aurora_findssrs/389ef73cf4e1/aurora_findssrs. Enter this path to the Path to a shed tool.

Everytime you update your XML file, you will need to integrate the updates by selecting the rmarkdown_report.xml and click Update XMLs in galaxy_tool_directory_folder.

Restart Galaxy instance

To have the updates in XML to take effect, you will also need to restart Galaxy instance. Use the following command to restart Galaxy

docker exec -it gtg_galaxy sh -c 'supervisorctl restart galaxy:'

You expect to see the following stdout.

galaxy:galaxy_nodejs_proxy: stopped
galaxy:handler0: stopped
galaxy:handler1: stopped
galaxy:galaxy_web: stopped
galaxy:galaxy_nodejs_proxy: started
galaxy:galaxy_web: started
galaxy:handler0: started
galaxy:handler1: started