Wondering about how laboratories out there that pursue wet lab work as well as bioinformatics centralize their computational protocols and pipelines. I am considering setting up such a centralized resource for my lab, where everyone doing data analysis can submit minute detailed aspects of their computational pipelines. Any advice or opinion on this matter is much appreciated thank you!
We have our own GitLab instance. One of the main issue is data management and for this, I am looking into iRODS. I use Rstudio notebooks to document data processing and analysis when using R, markdown files otherwise.
Interesting. I've been using GitHub for code and its markdown pages for documentation, similar to Ram's answer I guess. GitLab seems really interesting, do you have any particular reason to prefer it over GitHub?
About iRODS, I think/thought this was a pretty heavy weight solution, requiring a dedicated administrator, not really suitable to a small to medium size team. Also here, I'd like to hear any comment you may have...
For various reasons, we can't/don't want to rely on a third party so we have to have our own server. GitHub's source code is not available but GitLab's code is. By the way, we're now also running a mattermost instance with our GitLab server.
Concerning iRODS, the Sanger institute is part of the iRODS consortium and using it. See this paper for some details. I don't have experience with it yet but given what I've seen so far in terms of data management in various places, I think it is worth considering if you already have someone dedicated to running your IT or data management operations. Although there may be some time and effort involved at the beginning, I believe this should pay for itself in the medium to long term.
I guess in my question I am even looking for something simpler since I am the primary person in the lab doing bioinformatics. Just wanted a place to easily and quickly share workflows and pipelines.
Like others, we are now using GitLab, but before this we were using Trac's wiki, as it is easy to install and run, and easy to backup. While I strongly recommend to learn to use source code management systems, if your colleagues are just not yet there and you just need a place to cut-and-paste code and attach files to pages, Trac may be a good start. On Debian and Ubuntu systems: apt install trac.
We use bitbucket repos for reusable code, as well as one time scripts that were created for specific datasets. I personally maintain a lab notebook (in markdown) on a bitbucket repo as well.
Our team uses the institute's HPC, and we also have dedicated directories for code, resources and such. When I wear my Data Manager hat, I get to streamline backup/archive of gigantic raw and processed datasets as well, and the details go into extensive logs and also into my well-organized, searchable lab notebook :)
Interesting. I've been using GitHub for code and its markdown pages for documentation, similar to Ram's answer I guess. GitLab seems really interesting, do you have any particular reason to prefer it over GitHub? About iRODS, I think/thought this was a pretty heavy weight solution, requiring a dedicated administrator, not really suitable to a small to medium size team. Also here, I'd like to hear any comment you may have...
For various reasons, we can't/don't want to rely on a third party so we have to have our own server. GitHub's source code is not available but GitLab's code is. By the way, we're now also running a mattermost instance with our GitLab server. Concerning iRODS, the Sanger institute is part of the iRODS consortium and using it. See this paper for some details. I don't have experience with it yet but given what I've seen so far in terms of data management in various places, I think it is worth considering if you already have someone dedicated to running your IT or data management operations. Although there may be some time and effort involved at the beginning, I believe this should pay for itself in the medium to long term.
I guess in my question I am even looking for something simpler since I am the primary person in the lab doing bioinformatics. Just wanted a place to easily and quickly share workflows and pipelines.
Like others, we are now using GitLab, but before this we were using Trac's wiki, as it is easy to install and run, and easy to backup. While I strongly recommend to learn to use source code management systems, if your colleagues are just not yet there and you just need a place to cut-and-paste code and attach files to pages, Trac may be a good start. On Debian and Ubuntu systems:
apt install trac.