Multiple inputs in a workflow (variable number of inputs) ?
Hello, The number of inputs is not fixed for my workflow, maybe I have 6 inputs, maybe more than that, the number of inputs depends on what data I have for the analysis. As I know, in a workflow, all inputs should be pre-defined...that is to say, the number of inputs is known before running one workflow. Do you have any idea for this problem? How can I do to make a workflow with the variable number of inputs? In advance, thanks for the answer, Yufei
Very likely you have to use the API for these kinds of workflows. The UI for Galaxy is very limited in workflows that deviate from the One Input paradigm. Even for the most common workflow (paired end DNA sequencing of multiple samples ) you have to resort to using the API...Puzzling... Regards, Thon de Boer, Ph.D. Bioinformatics Guru +1-650-799-6839 thondeboer@me.com LinkedIn Profile On Nov 8, 2012, at 5:13 AM, LUO Yufei <yufei.luo@versailles.inra.fr> wrote:
Hello,
The number of inputs is not fixed for my workflow, maybe I have 6 inputs, maybe more than that, the number of inputs depends on what data I have for the analysis. As I know, in a workflow, all inputs should be pre-defined...that is to say, the number of inputs is known before running one workflow. Do you have any idea for this problem? How can I do to make a workflow with the variable number of inputs?
In advance, thanks for the answer, Yufei ___________________________________________________________ Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at:
The general case is hard and it was brought up repeatedly at the Galaxy community conference. One path might be this work (https://bitbucket.org/galaxyp/galaxy-central-homogeneous-composite-datatypes) on defining datasets that represent homongenous collections of files, but there are no easy answers currently. As for the comment about paired-end sequence workflows, such workflows should actually be doable now in Galaxy (since pull request 75). All you need to do is comment out line 40 of templates/workflow/run.mako and then you should be good to go. -John On Sat, Nov 10, 2012 at 4:04 PM, Anthonius deBoer <thondeboer@me.com> wrote:
Very likely you have to use the API for these kinds of workflows. The UI for Galaxy is very limited in workflows that deviate from the One Input paradigm.
Even for the most common workflow (paired end DNA sequencing of multiple samples ) you have to resort to using the API...Puzzling...
Regards,
Thon de Boer, Ph.D. Bioinformatics Guru +1-650-799-6839 thondeboer@me.com LinkedIn Profile
On Nov 8, 2012, at 5:13 AM, LUO Yufei <yufei.luo@versailles.inra.fr> wrote:
Hello,
The number of inputs is not fixed for my workflow, maybe I have 6 inputs, maybe more than that, the number of inputs depends on what data I have for the analysis. As I know, in a workflow, all inputs should be pre-defined...that is to say, the number of inputs is known before running one workflow. Do you have any idea for this problem? How can I do to make a workflow with the variable number of inputs?
In advance, thanks for the answer, Yufei ___________________________________________________________ Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at:
___________________________________________________________ Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at:
Thanks JOhn for the tip... I take it it was disabled since it is still in development? I tried it, but you have to manually match the pairs it looks like to me? In my API way of running paired end, I actively search for the matching read 2 for every read1 I get (_R1 and _R2 suffix)...We need to be able to match the filenames for files somehow (I know galaxy eschews the ideas of actual names for the output it creates, so probably not something galaxy would want to implement, but for automated workflows were I routinely get flowcells with 256 FASTQ PAIRS, manually matching which pairs should be matched, is not a viable option.... Using the names of files is the only option and maybe using a regular expression to match files would be the most generic implementation (but it would rely on names of the files being informative and not "2: FASTQ Groomer on data 1" etc. Regards, Thon de Boer, Ph.D. Bioinformatics Guru +1-650-799-6839 thondeboer@me.com LinkedIn Profile On Nov 12, 2012, at 6:29 AM, John Chilton <chilton@msi.umn.edu> wrote:
The general case is hard and it was brought up repeatedly at the Galaxy community conference. One path might be this work (https://bitbucket.org/galaxyp/galaxy-central-homogeneous-composite-datatypes) on defining datasets that represent homongenous collections of files, but there are no easy answers currently.
As for the comment about paired-end sequence workflows, such workflows should actually be doable now in Galaxy (since pull request 75). All you need to do is comment out line 40 of templates/workflow/run.mako and then you should be good to go.
-John
On Sat, Nov 10, 2012 at 4:04 PM, Anthonius deBoer <thondeboer@me.com> wrote:
Very likely you have to use the API for these kinds of workflows. The UI for Galaxy is very limited in workflows that deviate from the One Input paradigm.
Even for the most common workflow (paired end DNA sequencing of multiple samples ) you have to resort to using the API...Puzzling...
Regards,
Thon de Boer, Ph.D. Bioinformatics Guru +1-650-799-6839 thondeboer@me.com LinkedIn Profile
On Nov 8, 2012, at 5:13 AM, LUO Yufei <yufei.luo@versailles.inra.fr> wrote:
Hello,
The number of inputs is not fixed for my workflow, maybe I have 6 inputs, maybe more than that, the number of inputs depends on what data I have for the analysis. As I know, in a workflow, all inputs should be pre-defined...that is to say, the number of inputs is known before running one workflow. Do you have any idea for this problem? How can I do to make a workflow with the variable number of inputs?
In advance, thanks for the answer, Yufei ___________________________________________________________ Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at:
___________________________________________________________ Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at:
participants (3)
-
Anthonius deBoer -
John Chilton -
LUO Yufei