I appreciate your advice. This was what I have been looking for. I will
update the installer with it in the next release.

Thanks!

Jin

On Wed, Apr 6, 2016 at 8:20 AM, Ryan Dale [email protected] wrote:

bioconda https://bioconda.github.io/ might help make the software
installation much faster and reproducible. For example if you write a
requirements.txt that looks something like:

bwa ==0.7.3
samtools ==0.1.19
bedtools ==2.19
ucsc-wigtobigwig
ucsc-bedgraphtobigwig
ucsc-bigwiginfo
ucsc-bedclip
matplotlib
numpy
scipy

...and everything else you need

Then everything can be installed with:

conda install
--prefix $SOFTWARE_HOME/software_bds
--file requirements.txt
--channel bioconda

A couple details:

• it looks like in install_dependencies.sh you manually separate
  Python environments, which can also be handled by conda. You would need a
  second requirements file and would need to choose a different directory
• there are a couple of tools not yet on bioconda, but that's an easy
  fix. Exception to that might be MCR due to licensing
• --prefix $HOME/software_bds in the conda command creates the
  environment there, minimizing changes to the pipeline

Just a suggestion. I used to have installation scripts like this but then
found conda. Then bioconda came along and helped more. Made things much
easier for other people to get things running and much less brittle.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub
#1

from chipseq_pipeline.

AQUAS pipeline now supports anaconda virtual environment. You don't have to add those parameters if you have installed dependencies with install_dependencies.sh.

bds chipseq.bds -conda_env [CONDA_ENV_PY2] -conda_env3 [CONDA_ENV_PY3]

However, some R packages (snowfall, spp, ...) in R 2.15.x, Anshul's software (phantompeakqualtools and align2rawsignal) and MCR were not available in official conda channels. So you still need to install them with an updated install_dependencies.sh.

from chipseq_pipeline.

Very nice, thanks.

Just FYI, spp is available in the bioconda channel as well as some other R packages. I'd be happy to help get recipes for the other dependencies added as well if you're interested.

from chipseq_pipeline.

I don't think the latest version of SPP is on there. It (SPP) was updated
just a few months ago with changes we specifically added.

Anshul
On Apr 11, 2016 6:07 AM, "Ryan Dale" [email protected] wrote:

Very nice, thanks.

Just FYI, spp is available in the bioconda https://bioconda.github.io/
channel as well as some other R packages. I'd be happy to help get recipes
for the other dependencies added as well if you're interested.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub
#1 (comment)

from chipseq_pipeline.

It would be really great if you would help with missing dependencies.

Anshul
On Apr 11, 2016 6:07 AM, "Ryan Dale" [email protected] wrote:

Very nice, thanks.

Just FYI, spp is available in the bioconda https://bioconda.github.io/
channel as well as some other R packages. I'd be happy to help get recipes
for the other dependencies added as well if you're interested.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub
#1 (comment)

from chipseq_pipeline.

The latest version of SPP is here https://github.com/hms-dbmi/spp
On Apr 11, 2016 9:12 AM, "Anshul Kundaje" [email protected] wrote:

I don't think the latest version of SPP is on there. It (SPP) was updated
just a few months ago with changes we specifically added.

Anshul
On Apr 11, 2016 6:07 AM, "Ryan Dale" [email protected] wrote:

Very nice, thanks.

Just FYI, spp is available in the bioconda https://bioconda.github.io/
channel as well as some other R packages. I'd be happy to help get recipes
for the other dependencies added as well if you're interested.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub
#1 (comment)

from chipseq_pipeline.

Ah great, I missed that SPP is on github now. Good to know. I will look into getting the dependencies sorted out.

from chipseq_pipeline.