msconvert 1.3.post3

A wrapper for the pwiz/msconvert docker container with memory limits and automatic retry.

A Python wrapper for the pwiz/msconvert docker image.

Why?

In my personal experience, running the docker container directly may cause some problems:

• On large machines, the containers consume a lot of memory because the garbage collection rarely kicks in.

• Memory seems to leak when rescoring multiple files in one msconvert run.

• Limiting the memory usage is difficult because conversion may require somewhere between ~3G and ~30G.

• The conversion is single-threaded.

How to Run

Install the package via:

pip install msconvert

Pull the docker image:

docker pull chambm/pwiz-skyline-i-agree-to-the-vendor-licenses:latest

Run it via:

python -m msconvert

By default, the module will convert every *.raw file in the current directory to .mgf. However, there are other options available:

• -i: Input path (default is current directory)

• --input-format: Input format (default raw)

• --output-format: Output format (default mgf)

• --concurrency: Concurrency (default is the number of cores)

How it Works

• msconvert-py runs each conversion in a dedicated docker container.

• Each container is initially limited to 3G memory.

• When a container fails due to an out-of-memory kill, it will be re-run with an additional 8G memory.

• The containers run concurrently, by default up to the number of cores.

• Containers are only started when memory is available within the defined memory limit.