Direct RNA Polyadenylation Analysis

Running The Pipeline

Create a new virtual environment (python 2.7 recommended) and activate, either with virtualenv:

$ virtualenv polya-env
$ source polya-env/bin/activate

or conda:

$ conda create --name polya-env
$ source activate polya-env

The pipeline requires the albacore basecaller (we used version 2.3.3); set the BASECALL variable in the makefile to the location of your read_fast5_basecaller.py file as appropriate. Albacore is (as of pixel time) closed source but may be downloaded via the ONT community website.

Then, run

make -f pipeline.make

to download all datasets and code and generate all relevant files.

Notes

• It can take a long time to download and untar the fast5 files (a few hours to a full workday). Plan accordingly, and monitor progress with (e.g.) a watch ls -lah on the appropriate directories.

• This pipeline must be run on a server with an internet connection (to download the datasets from the European Nucleotide Archive). We recommend creating a fresh virtual environment and installing ONT's Albacore basecaller within that virtual environment before running the makefile pipeline. We recommend a server with 8 cores and at least 20GB of memory per core. We assume R, Rscript, and python are accessible from the command line and that the ggplot2 R package is pre-installed. We tested our pipeline with python 3.4.6, R 3.4.4, and ggplot2 v2.2.1.

• There is an additional dataset, 60xbpolyAb, which is commented out and excluded from the pipeline by default. This dataset is a second run of the 60xpolyA library and has nearly identical distributions/densities. It can easily be downloaded and included in the analysis by uncommenting the lines where appropriate.

Citations

Nanopore native RNA sequencing of a human poly(A) transcriptome, Rachael E Workman, Alison Tang, Paul S. Tang, Miten Jain, John R Tyson, Philip C Zuzarte, Timothy Gilpatrick, Roham Razaghi, Joshua Quick, Norah Sadowski, Nadine Holmes, Jaqueline Goes de Jesus, Karen Jones, Terrance P Snutch, Nicholas James Loman, Benedict Paten, Matthew W Loose, Jared T Simpson, Hugh E. Olsen, Angela N Brooks, Mark Akeson, Winston Timp, bioRxiv: https://www.biorxiv.org/content/early/2018/11/09/459529