Skip to content

Latest commit

 

History

History
68 lines (52 loc) · 2.59 KB

RepEnrich2_guide.md

File metadata and controls

68 lines (52 loc) · 2.59 KB
Raw
title description category subcategory tags
How to run repeat enrichment analysis
This guide shows how to run RepEnrich2
research
rnaseq
annotation

RepEnrich2 tries to look at something that standard RNA-seq pipelines miss, the enrichment of repeats in NGS data. It is extremely slow and is a pain to get going. Below is a guide getting it working and has some links to a fork of RepEnrich2 I made that makes it more friendly to use.

I have not actually validated the RepEnrich2 output, so caveat emptor.

Preparing RepEnrich2

Create isolated conda environment

conda create -c bioconda -n repenrich2 python=2.7 biopython bedtools samtools bowtie2 bcbio-nextgen

Download my fork of RepEnrich2

This has quality of life fixes such as memoization of outputs so if it fails you don't have to redo steps.

git clone git@github.com:nerettilab/RepEnrich2.git

Download a pre-created index

You can make your own, for example I made hg38 and the RepEnrich2 folks have mm9 and hg19 here. But the RepeatMasker file it uses needs to be cleaned first and I'm not sure how they cleaned it. They had a hg38 one cleaned already from RepEnrich so I just used that.

Download bcbio_RepEnrich2

Download bcbio_RepEnrich2. This will need modification if you want to use it, but it is simple, I just didn't bother as I don't anticipate us running this again.

Running RepEnrich2

bcbio_RepEnrich2 is all you need to run it, the help should give you enough information to go on. annotation here is the file from RepeatMasker that was used to generate the RepEnrich setup. The bowtie index is a bowtie2 index of the genome you aligned to. Running RepEnrich2 takes FOREVER, so be sure to run it on the long queue.

Example command:

python bcbio_RepEnrich2.py --threads 16 ../human-dsrna/config/human-dsrna.yaml /n/app/bcbio/biodata/genomes/Hsapiens/hg38/bowtie2/hg38 metadata/hg38_repeatmasker_clean.txt metadata/RepEnrich2_setup_hg38/

RepEnrich2 outputs

You will get three files for each sample, for example:

P1722_class_fraction_counts.txt
P1722_family_fraction_counts.txt
P1722_fraction_counts.txt

The class and family files are the counts in the samplename_fraction_counts.txt file aggregated by family or class. Those could be used as aggregate analyses, but the fraciton_counts looks at the different repeat types individually, so is more what folks are probably looking for.