We are working on a workflow project, created an element with command line that has 2 input port, however only 1 input port could be connected whereas the other cannot be connected at all.

Anyone has an idea what might be the problem?

Thanks!

---

This does looks like a bug.

Could you send us the element with 2 ports so we can reproduce and fix it?

Hi, we had tried with several solution and we had figured out that when we connect 1 read element to 2 output it works, however from the 2 output elements they could not converged back to a third element. Would this be due to the processing ability of the elements? Kindly advice, thanks! =)

When you create multiple outputs (for example from read-sequence) it means that the same data is passed to different elements. Once processed by the element new data context is created.

When you you want to move multiple data streams into the one element you actually need a data context multiplexing rule: tell to workflow how do you want to join the data. Examples:
1) for every element from port A wait for the first single element from port B
2) for every element of port A wait for every element from a port B
3) for every element from port A wait for the element from port  B with the same name/length/property

etc...

This becomes rather complex and for now only a very limited set of built-in elements can do such multiplexing. We're working on WD the architecture to support this feature in UGENE 1.11 version. The main problem for us is how to make it both simple and efficient.

For your task I would propose the following:
Now you have:
R->A1
R->A2
A1->E
A2->E

do not separate outputs from R into 2 streams. Use a single  pipe:
R->A1->A2->E
This will keep the data creation context and you will not need multiplexing to join different contexts.

In this case in A2 select  R-output but not A1-output to use.
For element E as input you will have all kind of outputs available: R-output, A1-output and A2-output so you can merge and use them as you wish.

Please tell if you need more details about this algorithm

Thanks for the guidelines on the workflow, it has been validated but new issue has arise: We can't integrate external tools (not in the UGENE package) into custom element by CMD tools. It kept giving out error of unknown element detected. Any idea what went wrong? Thanks again for the help! :)

kenkl wrote on Jan 20^th, 2012 at 4:09pm:

Thanks for the guidelines on the workflow, it has been validated but new issue has arise: We can't integrate external tools (not in the UGENE package) into custom element by CMD tools. It kept giving out error of unknown element detected. Any idea what went wrong? Thanks again for the help! :)

Hello. Another user of UGENE has found this problem too. See:
http://ugene.unipro.ru/forum/YaBB.pl?num=1326766085

We are working on it and it will be fixed at UGENE 1.10.2 version in a couple of weeks.

kenkl wrote on Jan 20^th, 2012 at 4:09pm:

Thanks for the guidelines on the workflow, it has been validated but new issue has arise: We can't integrate external tools (not in the UGENE package) into custom element by CMD tools. It kept giving out error of unknown element detected. Any idea what went wrong? Thanks again for the help! :)

We have resolved this issue. You can try it by downloading the latest snapshot from our site or waiting for the 1.10.2 version.

Thanks for your feedback.

Hi there,

About the algorithm on R > A1 > A2 > E, I had tried it on my elements but faced problem executing it, ended up with errors. Attached are the screenshots. Kindly advice.

Thanks!

---

As I can see from the first screenshot, you execution string of the left ttuner element is "ttuner -sa $out -if $in".
The ttuner key -if means, that its input parameter is the path to the text file with pathes to other abi files.

Are you sure that it is correct? Your "File list" produces the path exactly to such a text file or not?

You could use this element next way:
Input directory - path to the directory with ABI files. The directory, but not the file.
Absolute output paths - true
Recursive reading - true or false. You could read filenames only from the input directory or from every nested directory.
Include name filter - I think that it is *.ab1
Exclude name filter - if you don't want to produce some files, you could set up this parameter and exclude them.

So, your execution string would be without the -if key:
ttuner -sa $out $in

Yes, the path is correct as I had tested it individually and it works in the way of:

R > A1 > E and R > A2 > E

but doesn't work in the form of R > A1 > A2 > E

Not very sure why. I had tried your suggestion too but it still does not work. Kindly advice, thanks!

Kenneth

kenkl wrote on Feb 21^st, 2012 at 9:55am:

Yes, the path is correct as I had tested it individually and it works in the way of: R > A1 > E and R > A2 > E but doesn't work in the form of R > A1 > A2 > E Not very sure why. I had tried your suggestion too but it still does not work. Kindly advice, thanks! Kenneth

Yes, I can reproduce it too. It will be fixed soon.

kenkl wrote on Feb 21^st, 2012 at 9:55am:

Yes, the path is correct as I had tested it individually and it works in the way of: R > A1 > E and R > A2 > E but doesn't work in the form of R > A1 > A2 > E Not very sure why. I had tried your suggestion too but it still does not work. Kindly advice, thanks! Kenneth

Kenneth, your problem is fixed. The latest snapshot with the fix will be available tomorrow.

And a little advice for you. Set the output slot type of the right ttuner element as a text file, but not as a fasta sequence. So, the last element will be "Write plain text". It is needed because
"ttuner -qa $out $in"
produces not a fasta file, but some text qualities that are not supported by UGENE. So, if you use "Text type" for this slot then the data would be written as it is, without any changes.

---

Built snapshots are available for downloading.
http://ugene.unipro.ru/snapshot.html

Hi Sir, I had downloaded the latest snapshot and tried running my previous workflow on it but has error, not sure of why. I exactly the same as how it works before the snapshot today but couldn't work. (see screenshot 2)

As for the R > A1 > A2 > E, i tried, it ran but no output files are located. (See screenshot 1)

Kindly advice. Thanks!

---

Besides, i also felt confused when I run my workflow of:

File list > ttuner1 > E

Which works well in previous snapshot, but works awakwardly as it gave error notice on log but yet there is output file created. (See snapshot) This happens for other previous workflow as well.

---

I had deleted the config file in previous version, tried to run on 1579 snapshot but could not. However it could run on 1561 snapshot.

It looks like UGENE (ugene or ugeneui or ugenecl) doesn't have the writing permission to the temporary directory:
/tmp/ugene_tmp/

Try to grant them, please.

Hi there, I had granted full rights but still it does not work. See screenshot.

---

I can't reproduce this problem. I would be helpful if you share your files with me:
1) the scheme *.uwl file
2) custom workers *.etc files (~/.config/Unipro/ExternalToolConfig/*)
3) Maybe, some of input files, if their size is not too large.

You could send them to ugene@unipro.ru

Hi German, I'm not sure where does the *.etc file located, kindly advice. Thanks!

kenkl wrote on Feb 27^th, 2012 at 10:42am:

Hi German, I'm not sure where does the *.etc file located, kindly advice. Thanks!

You should enter your home directory ("/home/kenneth/", if you username is "kenneth"). Then follow this path: ".config/Unipro/ExternalToolConfig/". This folder contains your custom element files.

The ".config" filder could be hidden, so you need to show hidden files.

Here are the *.uwl and *.etc files:

---

Input files are too large to be attached. But you can try using PEH's file for testing, its similar.

http://ugene.unipro.ru/forum/YaBB.pl?num=1326766085/0

Thanks!

kenkl wrote on Feb 27^th, 2012 at 1:44pm:

Input files are too large to be attached. But you can try using PEH's file for testing, its similar. http://ugene.unipro.ru/forum/YaBB.pl?num=1326766085/0 Thanks!

Kenneth, you files are received. The problem will be investigated asap.

Kenneth, the problem is resolved.
The latest Linux 32-bit snapshot is available now. Other snapshots will be built and available in several time.