These are chat archives for nextflow-io/nextflow

11th
Dec 2014
Andrew Stewart
@andrewcstewart
Dec 11 2014 21:38
Heya. Any advise on setting up a system wide install of nextflow? I'm wondering if there are any considerations to factor in
like.. if I put nextflow into a shared area on a multi-user system, can all users see the same pipeline tracking info?
Andrew Stewart
@andrewcstewart
Dec 11 2014 22:00
looks like I can just set NFX_HOME to shared location
git credentials may be an issue though
Paolo Di Tommaso
@pditommaso
Dec 11 2014 22:41
Hi
Yes you can set a share folder as nextflow home
pipelines are downloaded /installed into the nextflow home
Andrew Stewart
@andrewcstewart
Dec 11 2014 22:45
thanks
Paolo Di Tommaso
@pditommaso
Dec 11 2014 22:45
however in my opinion it's better to have it in the user home
in this way each user is free to work with his/her pipelines
Andrew Stewart
@andrewcstewart
Dec 11 2014 22:46
Yeah, I can see the potential for conflicts
Paolo Di Tommaso
@pditommaso
Dec 11 2014 22:46
yes, at the end the real share storage is github
(or bitbucket)
Andrew Stewart
@andrewcstewart
Dec 11 2014 22:47
Im really liking cextflow btw
gah, autocorrect
Paolo Di Tommaso
@pditommaso
Dec 11 2014 22:47
:)
Andrew Stewart
@andrewcstewart
Dec 11 2014 22:47
every time I hit the docs I discover something new. This is so feature rich.
Paolo Di Tommaso
@pditommaso
Dec 11 2014 22:48
yeah, there's really a lot of work
new features are coming next week
Andrew Stewart
@andrewcstewart
Dec 11 2014 22:48
my highest praise to everyone involved
Paolo Di Tommaso
@pditommaso
Dec 11 2014 22:48
thanks
spread the word and tweet/blog about it, if you can
Andrew Stewart
@andrewcstewart
Dec 11 2014 22:49
I ran parallel jobs via SGE inside docker containers. That's amazing.
Paolo Di Tommaso
@pditommaso
Dec 11 2014 22:49
Yes, that's very cool, I would really like about your use case,
take in consideration to blog about that
I mean .. "I really would like to know about your use case"
my "autocorrect" is broken as well ;)
Andrew Stewart
@andrewcstewart
Dec 11 2014 22:52
I can't figure out how to turn it off in yosemite
Paolo Di Tommaso
@pditommaso
Dec 11 2014 22:52
ah, I refused to upgrade to it ..
Andrew Stewart
@andrewcstewart
Dec 11 2014 22:53
But yeah, I want to try to open source some RNASeq pipelines my group is developing with nextflow
Paolo Di Tommaso
@pditommaso
Dec 11 2014 22:53
may I ask you what/where is your group?
Andrew Stewart
@andrewcstewart
Dec 11 2014 22:54
Sure. Small/medium sized biotech based out of South San Fran: http://www.veracyte.com
Paolo Di Tommaso
@pditommaso
Dec 11 2014 22:55
wow cool
Andrew Stewart
@andrewcstewart
Dec 11 2014 22:55
We develop medical diagnostics
Paolo Di Tommaso
@pditommaso
Dec 11 2014 22:56
cost to cost, tech transfer from Barcelona to SF :)
Andrew Stewart
@andrewcstewart
Dec 11 2014 22:56
So we're primarily developing our pipeline for feature engineering in classification
Paolo Di Tommaso
@pditommaso
Dec 11 2014 22:57
I see, rna-seq data analysis..
Andrew Stewart
@andrewcstewart
Dec 11 2014 22:57
Ah, Visca Catalunya
I used to have a bunch of roommates from Barca :)
Paolo Di Tommaso
@pditommaso
Dec 11 2014 22:57
Great
Andrew Stewart
@andrewcstewart
Dec 11 2014 23:29
I have a question
Paolo Di Tommaso
@pditommaso
Dec 11 2014 23:29
sure
Andrew Stewart
@andrewcstewart
Dec 11 2014 23:30
how come it seems like an input is often fed into file() twice
i.e. line 37
and then line 44
or is that just like a 'type declaration' or something on 44
im not that familiar with groovy
Paolo Di Tommaso
@pditommaso
Dec 11 2014 23:31
ok, the file(x) method (line 37) is required to convert the parameter string to a file object
instead the file in the input block is required to "qualify" the input
so that it is able to stage that file in the process working directory
Andrew Stewart
@andrewcstewart
Dec 11 2014 23:34
ok
so its basically just asserting the fact that its a file object ?
Paolo Di Tommaso
@pditommaso
Dec 11 2014 23:34
yes
Andrew Stewart
@andrewcstewart
Dec 11 2014 23:34
got it, thanks!
Paolo Di Tommaso
@pditommaso
Dec 11 2014 23:36
the important thing is that when you declare an input as file (line line 44)
it tries to convert any value to a file
so at the end you can also feed it with a string value, and it will be converted automatically to a file, so that it can be used by the tool in the process script
For example
line 27
splits a fasta file in single sequences (strings)
then when is used as input for the file at line 36, it converts that fasta string to a file
so that it can be used by the script
Paolo Di Tommaso
@pditommaso
Dec 11 2014 23:44
Ok, I need to leave
Happy to help you, if you have more doubts, you may want to post in the google group
Cheers
Andrew Stewart
@andrewcstewart
Dec 11 2014 23:55
thanks again!