Tutorials

After logging in, you will be redirected to the experiments overview. If you are a new user, this page will be empty (Figure 3). Please note this page can list two types of experiments:

• Current experiments (experiments you uploaded and own),
• Shared experiments (experiments uploaded by others you are allowed to view)

Clicking on the add new experiment button ( icon) opens the experiment creation window (Figure 4).

To start, like shown in Figure 4, enter a name and a description for the experiment. For instance, Tutorial 1 as a name and Documentation tutorial 1 (Panicum) as description. Select PLAZA 4.5 monocots as reference database, as PLAZA is the recommended database for plants and microbial photosynthetic eukaryotes, and the 'monocots' version contains genomes from closely related species. In case data from other lineages is analyzed, we recommend selecting eggNOG 4.5. An overview of the available reference databases can be found on the tools & parameters documentation page.

Add the experiment by clicking create experiment.

The new experiment will now appear in the current experiments table (Figure 5). Note that each user can create a maximum of 20 experiments.

To continue, click on the experiment's name (e.g. Tutorial 1) in the current experiments table. This will take you to your newly created TRAPID experiment and display the experiment overview page (Figure 6). After transcripts sequences are processed, this page will contain general experiment statistics (in the Experiment information panel) and a detailed overview of the transcripts.

Two main elements may be used for navigating a TRAPID experiment: the experiment header (top) and the side menu (left). The experiment header contains links to experiment controls (jobs, log, and settings) and a search box to find specific sequences, gene/RNA families, or functional data. The side menu lists the main data import/export, exploration, and analysis options available within TRAPID. Note that in figure 6, all the items except Import data are disabled, since the experiment is empty.

To continue and upload input sequences, click on the Import data item of the side menu.

Input sequences must be provided in FASTA format. Make sure each sequence has a unique identifier (max. 100 characters) and no empty sequences are present. Please note that individual files cannot be larger than 32Mb. For large datasets, we therefore recommend to supply compressed data (zipped or gzipped) and/or to split the data in multiple files.

In case you downloaded the tutorial dataset, you can upload the file (click Browse... and locate it on your system). Otherwise, you can directly supply the URL (https://ftp.psb.ugent.be/pub/trapid/datasets/panicum/panicum_transcripts.zip). After selecting a file or URL, click Add file/URL to confirm your choice and add the dataset. Afterwards, click Load data into database to load all sequences into our database. Both steps are essential before the data can be processed.

After the sequences have been loaded into the TRAPID database, the next step is to perform the initial processing. This step will add them to gene or RNA families, add annotations and a taxonomic label (if this step was enabled). To start the initial processing, go to your TRAPID experiment. Note how the page now contains a table with transcripts at the bottom, and how some options of the side menu have become enabled (although they only become useful after the processing). Click Process transcripts.

On the next page, you have to specify how transcripts should be processed by TRAPID. For this tutorial, select the settings as shown in Figure 8 (default settings except for the taxonomic classification that was disabled). More details on these settings can be found in the general documentation. Finish by clicking Run initial processing. Depending on the size of the dataset, the selected settings, and the load on our servers, this can take up to several hours. For this tutorial, it should take around one hour after starting. In the meantime, the experiment will be in processing state and cannot be accessed, except for the experiment's job management and log pages. An e-mail will be sent upon completion.

Once the initial processing has finished, all sequences will be included in gene/RNA families and be annotated (when possible). Additionally, as transcript data often includes truncated sequences or sequences with indels, potentially problematic sequences are flagged.

Once the initial processing has finished, go to the experiment overview page. More options are now accessible from the side menu, and the table at the bottom of the page lists your sequences with their assigned gene family, predicted functional annotation, and meta-annotation. Click the Statistics > General statistics item in the side menu.

The following page will show a broad range of statistics that reveal the quality of the input dataset, how many sequences were assigned to gene or RNA families, and how many were functionally annotated (Figure 9). The other page available under Statistics, Sequence length distribution, displays the length distribution of the experiment's transcript and predicted ORF sequences.

Gene families (which group protein-coding genes derived from a common ancestor) and RNA families (which group homologs of known non-coding RNAs) are available from the side menu, under Browse gene families and Browse RNA families, respectively.

Relevant families can be found using the search function of the experiment header. For instance, by selecting GO term, relevant GO identifiers or descriptions labels can be searched, and the associated sequences retrieved.

First, search for leaf senescence GO terms, and look at the transcripts annotated with the leaf senescence (GO:0010150) term by clicking on the GO identifier (Figure 10).

Look at contig04501: this transcript is annotated as 'quasi full length'. Click on its identifier to get to the transcript page. Once there, click the associated gene family identifier, which will take you to its associated gene family page.

From the gene family page, build the multiple sequence alignment (MSA) for this gene family by clicking Create multiple sequence alignment / phylogenetic tree in the toolbox. On the next page, check Generate MSA only to ensure no phylogenetic tree is generated. While it's possible to use MUSCLE instead of MAFFT and to modify the set of input sequences, we will not do so in the tutorial. Click Create MSA/tree to launch the creation of the MSA. Note that if more than 250 genes are selected, it will not be possible to submit the MSA creation job.

After completion, the page will contain a tab with a viewer for the MSA. From the alignment, it can be seen the transcript has indeed a good alignment with most of the other members in the family, although at the N-terminal end there likely is a portion missing (and hence is indeed quasi full length). The other transcript assigned to the family, contig20276, is the opposite: the C-terminal end appears to be missing. Both contigs may therefore represent a single, split transcript.

The Files & extra tab provides a link to download the MSA in .faln format and lists the parameters used to generate it.

Now, search for another transcript: contig01069. Find the gene family page of this transcript and click Create multiple sequence alignment / phylogenetic tree. This time, we will create a phylogenetic tree. In addition to the settings mentioned in the previous part, various phylogenetic tree creation settings can be adjusted: the MSA editing mode, the tree construction algorithm, and tree annotations (extra information about the transcripts displayed next to tree leaves). Feel free to do so (for instance, selecting fewer species from the reference gene family).

Next, click Create MSA / tree. The tree creation job will be launched and an e-mail will be sent upon completion. Once the job has finished, go back to the gene family page and click View or create multiple sequence alignment / phylogenetic tree to view the tree. The page will now contain a tab with an interactive viewer for the generated phylogenetic tree (Figure 11).

In case no tree is displayed or get an error message, very often the selected MSA editing setting is too stringent and should be modified. To check if this is the case, go to the multiple sequence alignment tab, and check the length of the edited alignment: if it is zero amino acids, then the editing was too stringent.

The Files & extra tab provides links to download the tree in PhyloXML and Newick formats and lists the parameters used to generate the phylogenetic tree. In case you want to create a new MSA or tree for the gene family with different settings, simply click the Create MSA / phylogenetic tree […] link on top of the page.

Read more about multiple sequence alignments and phylogenetic trees in the general documentation

Within a TRAPID experiment, transcript subsets can be defined from any arbitrary list of transcript identifiers, for instance transcripts expressed in a specific condition or tissue, and used to perform subsequent within-transcriptome functional analyses. Subsets may either be uploaded as a file or created interactively from the web application. By creating transcript subsets, several new analyses become available, such as comparison of functional annotation between different subsets, or functional enrichment analysis.

In this tutorial, we provide an example set of 33 Cell Cycle transcripts. The list can be downloaded from TRAPID’s FTP. To create a new transcript subset, click Import data (side menu). In the Transcript subset tab, click Browse... to select the downloaded file and enter a name for the subset (e.g. Cell_cycle). Click Import subset to finish.

Transcripts subsets can be inspected and deleted from the Explore subsets page (side menu). To check if the cell cycle set is enriched for specific GO terms or Protein domains, click the Subset enrichment item in the side menu. On the next page, select a type of functional annotation for which to perform the analysis, a transcript subset, and a maximum q-value (corrected p-value) threshold. For this tutorial, we selected InterPro domains, Cell_cycle and 0.005. Click Compute enrichment to launch the analysis.

Figure 12 shows the resulting page displaying for each of the enriched InterPro domains the enrichment fold, significance and a short description. Note that the InterPro domain identifiers are hyperlinks to pages containing more detailed information, and that results may also be explored as a table or downloaded.

It is optionally possible to precompute functional enrichments for all available types of functional annotation, subsets, and q-value thresholds, by clicking the Run functional enrichment button on the experiment overview page.

Read more about subsets/labels and functional enrichment in the general documentation