bowtie-pp2.xmlconfiguration file has been imported, you can click the CLC bio Bowtie Map header to see the configuration as shown figure 12.13.
From an end-user perspective, when the configuration on the CLC Server is complete, they will be able to launch the CLC bio Bowtie Map tool via their Workbench Toolbox. A wizard will appear, within which they will select the sequencing reads to be mapped, identify the pre-built index file of the reference sequence to use and set a few parameters. The bowtie executable will then be executed on the server system and the results generated will be imported into the CLC Server using post processing tools. The sam mapping file is imported using the Import SAM/BAM Mapping Files tool. A fasta file of sequences mapping to multiple locations is imported using the Fasta High-Throughput Sequencing Import tool.
Below, we step through the General configuration panel and then explain the configuration of the post processing tools that handle the outputs from the bowtie analysis.
Each of the parameters (items within curly brackets) written into the "Command line" box is presented as an item in the General configuration panel. There, we define the type of information each parameter expects or represents and default values, where relevant.
To understand how these parameters relate to the information that will be passed to the native bowtie executable, please refer to the bowtie_map.sh script in the clcbio folder that should now be in place in the bowtie distribution folder.
Stepping through the parameters in the order they appear in the Command line area of the configuration, and thus the order they appear in the General config panel:
- The reads parameter refers to the data that will be provided to bowtie to map.
The User-selected input data option means the user will be able to select data in a CLC File or Data Location. This data will be exported from the CLC File or Data Location such that the bowtie tool can use it. The second element in this line specifies the format the data should be exported in. This is set to FASTA (.fa/.fsa/.fasta) as this is the format the bowtie tool expects sequencing read data.
- The index parameter is expecting the name of a bowtie index. Specifying the type Text for this parameter means a user will see a box in the Workbench Wizard that they can type text into. Here, a default name, "coli" has been specified, which can be changed by a user launching CLC bio Bowtie Map.
When setting up a tool like this, it would be simpler for users, and much less subject to error, if the type CSV enum were selected, and a specified set of indices were listed. Then, a drop down list of options would be provided to the user in a Workbench Wizard, when launching the external application, rather than relying on users typing in the correct names of available bowtie indices.
- The sam file parameter refers to the sam mapping file that bowtie will generate as one of its results file. Thus, the type is set to Output file from CL. Import of sam files into the CLC Server involves a tool that requires user input. Thus, a post processing tool is configured. This can be seen immediately by the text in the second drop-down box: "Linked with Import SAM/BAM Mapping Files".
If a parameter with type Output file from CL is not mapped to a parameter of a post processing step, the text displayed is "Do not standard import / map to high-throughput sequencing importer". Mapping of outputs to post processing tools is described in more detail below.
The last entry in the configuration of the sam file parameter is the name of the sam output file that bowtie should generate. Here it is set to sam_output. This file name is used by the bowtie command. The Workbench or Command Line Tools user never sees it.
- The max number of multimatches parameter allows a user to select the maximum number of locations a read can map equally well to for it to be included in the mapping. The type is set to CSV enum, which means a user will be able to select a value from a drop down list of the 3 values listed in the last field (2,3,4). The first value will be the default. The values in the middle field are those passed to the bowtie wrapper script and then onto bowtie. So, for example, if "2,3,4" were entered in the middle field, and "two, three, four" in the last field, a user could select the option "two", and bowtie would be sent the value 2.
- The multimatch filename parameter refers to another output from bowtie, this one containing fasta formatted reads that match to multiple locations of the reference equally well. Since it is a result file, the type is set to Output file from CL. We have decided to use a post processing tool to bring the results back into the CLC Server, the Fasta High-Throughput Sequencing Import tool.
- The max number of mismatches parameter allows a user to select the maximum number of mismatches to be allowed between a read and the reference in order for a read to be considered as matching the reference at that location. The type is set to CSV enum, and is presented to a user in the same way as the max number of multimatches parameter described above.
- The report all matches option is one that can be turned on or off. Thus it is set to type Boolean text. A user will be presented with a checkbox they can select or deselect in the Workbench Wizard. The value in the text field, here "-a", is the one bowtie will be passed if the user selects the checkbox. If the user does not select the checkbox, this parameter will not be sent to bowtie.
If you expand click on the High-throughput sequencing import /Post-processing link below the General configuration area, you will see that there are two post-processing tools selected: the Import SAM/BAM Mapping Files tool and the Fasta High-Throughput Sequencing Import tool.
In each case, clicking on the Edit and map paramaters button below it will bring up the configuration window for that tool. Here, several types of configuration can be carried out.
- Mapping of outputs of the external application to inputs of the post processing tool.
- Locking or unlocking of parameters, determining which parameters users can alter when launching the tool via the Workbench or Command Line Tools.
- Setting default values for parameters of the external application.
Here, we step through the configuration of the Import SAM/BAM Mapping Files tool. The configuration of the Fasta High-Throughput Sequencing Import is similar.
The parameters in this configuration window are the Import SAM/BAM Mapping Files tool parameters, just as would be offered when that tool is launched directly in a CLC Workbench.
A locked lock symbol by a parameter means that the user will not be given access to this option when launching the tool. Default settings for lock parameters are used. The locked parameters shown in figure 12.14 indicate that a track will be output rather than a stand-alone read mapping, unmapped reads will be saved, references will not be downloaded from an external source and, had they been, downloaded references will not be saved. Quality scores and sequence names will be kept (not discarded).
By contrast, the References parameter is unlocked. When using the Import SAM/BAM Mapping Files tool, users need to specify where the relevant reference sequences are. Thus, this option should be made available for users to configure when the tool is being launched.
The input to the Import SAM/BAM Mapping Files also needs to be defined. This is done by mapping the relevant output from the bowtie command to the input parameter for the Import SAM/BAM Mapping Files tool. The output from bowtie is defined by the "sam file" parameter, and the relevant input parameter in the import tool is "Selected files". A drop down list of potentially relevant parameters appears to the left of the "Selected files" parameter. In our example, this has already been mapped to the "sam file" parameter of the command, as shown in figure 12.14.
Note: The drop-down lists of possibly relevant parameters provided in the post processing tool configuration window are populated based only on the types of parameters (in the General config pane). Any parameters of a type that could be relevant are presented. This means that some parameters appearing in these lists may not make sense contextually.