First time use:
Make sure you have snakemake, hisat2, samtools, deeptools and subread installed, or are using a conda virtual environment with these installed.  Running 

$ conda env create -f envs/rnaseq-basic.yml

should create an environment with the right packages installed.  My conda environment is called rnaseq-basic, so before running the workflow I run:

$ conda activate rnaseq-basic

Then I do a practise run or "dry" run:

$ snakemake -s scripts/rnaseq_with_replicates.py --configfile test_mar19.json -npr

$ snakemake -s scripts/rnaseq2.py --configfile test_mar19.json -npr

scripts/rnaseq2.py runs the "proper" way to make hisat indices which leads to better alignment but takes up a lot of memory. For ensembl I needed to use laminB and 330GB of RAM for this step. 

scripts/rnaseq_with_replicates.py runs the quick and dirty way to make hisat indices.  

Full usage:

$ snakemake -s scripts/rnaseq_with_replicates.py --configfile test_mar19.json -pr

$ snakemake -s scripts/rnaseq2.py --configfile test_mar19.json -pr

Correct index creation in scripts/rnaseq2.py.  For ensembl I needed to use laminB and 330GB of RAM for this step.  

$ snakemake -s scripts/rnaseq2.py --configfile rnaseq_mar19.json --cores 8 -pr build_hisat_index > build_logs/02.log 2>&1 &