Running Trinity and Trinotate on a sample dataset:
Looks like I got Trinity and Trinotate to run, mostly, on a small sample data set provided by the developers. RNAMMER isn’t working, so I still have to fix that, but it’s progress!
R notebook: here
Results file: results.txt
Installing Trinity and Trinotate on Emu:
Looks like we’re getting some flounder RNA-seq data sometime soon, so Steven asked me to install and play with the Trinity and Trinotate pipeline. Install order/issues will be documented below. Some programs were already installed at the time of this writing, but I documented as though it was a complete install.
1. Install Trinity from here.
2. Install Jellyfish from here.
3. Install Trimmomatic from here.
4. Install Trinotate from here.
5. Install HMMER from here. Note: This is a different site than listed on the Trinotate installation instructions, but their site was down at the writing of this. Hopefully this one will work.
6. Install TransDecoder from here.
7. Install SQLite via
sudo apt-get install SQLite.
8. Install NCBI-BLAST+ from here.
9. Install optional programs signalP, tmhmm, and RNAMMER found at this site. Note: Each program requires email permission by the developer which is only good for 4 hours. Kind of a pain, but oh well. RNAMMER also requires some hacking, which is described in detail on the Trinotate website.
Building the Trinotate protein database:
sudo ./admin/Build_Trinotate_Boilerplate_SQLite_db.pl Trinotate is how you’re supposed to be able to build the custom SwissProt and PFam database required by Trinotate, but there were a few missing perl modules. So prior to running this,
sudo perl -MCPAN -e 'install DBD::SQLite',
sudo perl -MCPAN -e 'install Bundle::DBI'
Everything looks like it installed ok, and after adding everything to the $PATH in
/etc/environment, rebooting, and realizing I’d forgotten to save this midway through the reboot (thankfully it saved). I believe everything is done!
Next step, testing.