PBJelly pt 2.

So I have PBJelly working! After trying a fork of the PBJelly project, which fixed the blasr compatibility issue but introduced other issues I went back to the original developer’s version and installed a legacy version.

Installation of legacy version of Blasr:

git clone git@github.com:PacificBiosciences/pitchfork.git pitchfork
git checkout legacy_blasr
cd ~/pitchfork
make init
make blasr 
source ~/pitchfork/deploymentssetup-env.sh

This takes a while

Next, we have to install networkx. This was tricky due to permission issues with pip being unwilling to install a package without sudo, but that resulting in borked permissions for my local python install.

I came across python -m pip install networkx as a method to install relative to a specific python install, and all seemed well after that.

First, I run the setup script for PBJelly via source ~/Documents/PBSuite2/PBSuite_15.8.24/setup.sh to set $PATH stuff for running. You have to do this every time unless you want to add this to your bash profile to auto-execute.

Then, I test the Jelly install.

Then I get to working my way though the different steps.

Setup:

Mapping:

Support:

Extraction:

Assembly:

and finally, Output:

Which results in a bunch of files, most of interest is the .fasta file with our newly gap filled sequence

Looks like everything worked! Now I just need to get some spare time on Emu to run the O. lurida stuff. Hopefully soon. Or maybe on Hyak if the new emails are to believed.