caleblareau / proatac

Preprocessing pipeline for (sc)ATAC data

This package is maintained by Caleb Lareau in the Buenrostro Lab. Source code is made freely available and a packaged install version is provided through PyPi.

About

The proatac package implements our data processing and quality control pipeline bulk ATAC-Seq, single cell ATAC-Seq, and droplet-based ATAC-Seq data.

Workflow Overview

A detailed description of the workflow including a description of various parameter settings is discussed in depth in the proatac documentation. Below is a brief overview of

Installation/Documentation/FAQ/.etc

Check out the proatac documentation for detailed installation instructions, dependency configuration, and other

Questions/comments/feedback

are always welcomed. The easiest way for us to have correspondence (if appropriate/interesting for the public) is through raising a new issue on the GitHub source. For private concerns, email Caleb.

proatac logo made freely with logomakr these blue and red color definitions.

proatac's Issues

fastq permissions

Need to actually verify read permissions for each input fastq file

bigwig files

https://github.com/taoliu/MACS/wiki/Build-Signal-Track#Fix_the_bedGraph_and_convert_them_to_bigWig_files

Probably just use Rtracklayer instead of doing anything external

Docker

I should do one. Need both python 2.7 and 3.6 and R as languages in the base image

Python versions

This will be a Python 3 package by virtue of SNAKEMAKE being a Python 3 package

However, macs2 only runs in Python 2, so I'll have to create multiple instances of virtual environments in all likelihood to support an install from first principles

Sample table verification

Need to pass all fastq files through that are supposed to be interpreted

This will check for mis-specified files and also those that can't be read (bad permissions; see other issue)

proatac docs not rendering properly?

The generated pdf of the docs leaves out a few chapers (e.g alignment, chapter 5). This might be only a 'work in progress' bug, but I'll open this issue in case it is a pandoc / generation error.

see if there are actually fastq files in sequencing dir

I ran this thing where I told the .yaml that there were files in fastq/ but they were actually in fastq/stuff.

Recursive? Throw some sort of error? idk man but the current output was pretty rough.

proatac logo

MultiQC desires an absolute path to the logo but this is hard for an installable file path. Can copy the logo to user output but it isn't clear how to get it to get it to the report

Collapse multiple sample fastqs

Currently throws an error if detected; zcat them?

pigz the mito compression

currently just using gzip

Species mix

need to awk cobble together species mixing reference genome stuff for annotation/TSS and then find an FTP server for the combined reference genome