Frequently Encountered Problems

Welcome to the frequently encountered problems pages. Here we will list those problems that are often encountered by our users when running the UEA small RNA Workbench and how to solve them. Or solutions to tricky problems that have been solved by our users.

If you have encountered an issue and resolved it please inform our team so that we can post the information on this page. In addition, any issues that are solved numerous times through our comment or ticket system will be placed here.

If your problem is not listed here please feel free to contact us as usual by leaving a comment on any page.

We will continue to update these pages as more solutions are required

Version 4+ Specific: 

Known issues

  1. Display of file hierarchy may not work on the latest version of java. Workaround by right clicking and refreshing the page to make the hierarchy appear or use slightly older version of JRE (update 51 or below)
  2. Some differentially expressed sequences can be annotated but the actual name of the annotation may not appear. This is due to poorly formatted GFF files mostly. A fix is in progress and will appear in the next release
  3. Switching to plot view can often prevent the differential expression table from re-displaying. This is a fairly major issue and is currently under investigation…

No graphics displayed

This is a known issue relating to java. More specifically any version of the JRE/JDK after update 51. Every now and then it refuses to reload information entered into external JSON files (the files containing the data for plots and other graphics). A short term fix is to simply right click on the problem page and hit reload

Non local (server-client) X11 forwarding shows no graphics

This is a known problem, for some reason X11 has decided to render no interface when running from a server. However, X windows displayed directly from a linux desktop is unaffected. Investigations are ongoing…

Poor rendering performance for both the interface and graphs

The interface works far better in JRE version 1.8 update 51 compared to the latest build, therefore we strongly recommend using this build only for the time being. Unfortunately we are at the mercy of the java developers at this point as the interface runs very smoothly in the javascript engines that come packaged with popular web browsers such as Safari or Chrome.

The plot views that are visible from the first and second reports can be opened from a web browser if they become unusable. Open the export menu for the option.

All Versions:

OSX asks to install a legacy version of the JRE (version 6) to run the .app file from double clicking

This is due to the previous versions of Java being built by apple instead of Oracle (or Sun Microsystems at the time). DO NOT INSTALL JRE6! This will not allow the software to run. This is an issue you may encounter with any java software bundled into a .app file. To fix this, first enter this command:

sudo nano `/usr/libexec/java_home -V`/../info.plist

Then locate the lines:

<key>JVMCapabilities</key>
 <array>
  <string>CommandLine</string>
 </array> 

And replace them with:

<key>JVMCapabilities</key>
 <array>
  <string>JNI</string>
  <string>BundledApp</string>
  <string>WebStart</string>
  <string>Applets</string>
  <string>CommandLine</string>
 </array>

When testing this my machine also required a restart before this took effect. After that it should load as normal using the correct JRE/JDK. Please comment below if this does not resolve your issue.

OSX Mavericks and Mountain Lion show the .app file as being damaged:

This is because of the permissions system in the new versions of OSX. All applications downloaded from the internet must be shown as trusted or signed by an official apple developer and gained from the app store. To allow the workbench to run you can either change your security permissions in the control panel or open the package up (right click and show contents) navigate to the java folder and run the sRNAWorkbenchStartup.jar file by right clicking and selecting ‘open’ then tell OSX to allow this program to run.

A new version of the .app file will be released shortly that will negate this problem and allow the software to launch as normal

Since updating OSX the terminal has defaulted to OpenJDK and the Workbench does not fully support this build:

This is because Apple no longer support java. However, assuming you have the oracle version of java on your system you can select the official build of Java using the Java Preferences program (type Java Preferences into spotlight), alternatively you must add the location of the JRE/JDK to your path variable that is found within your bash profile. Something similar to this with the version number (shown in bold) set to match whatever you have on your system:

export PATH=/Library/Java/JavaVirtualMachines/1.6.0_29-b11-402.jdk/Contents/Commands:$PATH

I am quite certain that running the application from the jar or from the new launcher added in Version 3.0 will select the Oracle version if it is installed.

I am getting an error similar to the following in the console while trying to run miRCat (RESOLVED):

SEVERE: WORKBENCH: MIRCAT: Message: Index: 0, Size: 0;
 Stack Trace: java.util.ArrayList.rangeCheck(ArrayList.java:604)

or

SEVERE: WORKBENCH:
 java.net.ConnectException: Connection refused

This is likely due to some kind of premature close of the workbench possibly from a unrecoverable crash leaving zombie threads hanging. Until a solution presents itself in an upcoming fix, this can be resolved by ensuring all java processes related to the workbench are closed before attempting to start miRCat again.

I am having trouble using X11 forwarding through a PuTTY client on Windows 7

Install MobaXterm as X server. http://mobaxterm.mobatek.net/support/documentation.html#4_2_5 “I have an issue with an X11 remote program (Java/X11/Motif) which does not accept keyboard input: everytime I press a keyboard key, nothing is written on the text field

This problem can generally be solved easily by using X11 with a window manager instead of using X11 in “multiwindow” (transparent) mode: go to MobaXterm “Settings” window –> “X11″ tab –> select “Windowed mode with Fvwm” setting in the combo box and apply. Restart your remote program and it should take keyboard input correctly.
If you want to use the standard “multiwindow” (transparent) X11 server and occasionally use the windowed X11 server with the window manager, you can go to MobaXterm “Settings” window –> “X11″ tab –> select “Multiwindow mode” setting in the combo box and apply, then open the “tools” menu and select either “X11 tab with Dwm” or “X11 window with Fvwm2″.

(thanks to Marius Snyman for the fix)

I am getting no output from any tools on a 64bit LINUX based operating system or an error similar to:

PATMAN ERROR: Patman result file was not created:

Have you ensured the 32bit libraries are installed? An example of how to do this on the Ubuntu operating system would be:

sudo apt-get install ia32-libs

I am getting an out of memory warning from one of the tools in the workbench:

The startup program will only allocate a portion of RAM it considers as being “available” to the operating system. First you must ensure that there are no other memory hungry applications running before launching the Workbench. If you are using the system on OSX you can also run the ‘purge’ command at the terminal to free any inactive memory.

Alternatively, if you feel the workbench can use all of your available memory you can allocate RAM yourself using the -Xmx -Xms commands at the command prompt.

After navigating your terminal to the location of the workbench files you can execute the following command (where N is the amount of memory in GB you wish to use):

java -XmsNg -XmxNg -jar Workbench.jar
 
  • Franklin

    Hi , there is no response whatever i click continue workflow after File Review. It is still waiting for user input and circle in yellow. I am using mac, java 1.8.0_45

    Cheers,
    Franklin

    • The sRNA Workbench

      Hi Franklin,

      Was there any errors written into the console? Can you try with the latest java and build of the workbench?

      Cheers,
      Matt

  • Kamya Singh

    I am facing problem in running the jar files. the work bench file is not getting executed. Can some one please help me solve this problem.

    • The sRNA Workbench

      Hi Kamya,

      Can you give a little more information about the problem you face?

      Can you tell me what OS you are using, which java version and what happened when you tried to run the software?

      Cheers,
      Matt

  • Pingback: The small RNA Workbench Version 4.0 ALPHA | The UEA small RNA Workbench()

  • Rob King

    What format do the input fasta files have to be? I’ve tried mirdeep2 and have the format as below but getting no results using mircat.

    >seq_0000005_x1749581
    AAACTTTCAACAACGGATCTC

    Looking at the tutorial files looks like need to convert to below?

    >seq_0000005 (1749581)
    AAACTTTCAACAACGGATCTC

    • The sRNA Workbench

      Hi Rob,

      The format is a slightly redundant one, a hang over from the original tools web based programs. I will be changing this at some point, but it works as follows:

      >SEQUENCE(ABUNDANCE)
      SEQUENCE

      for example:
      >ACTG(10)
      ACTG

      filter can produce the format for you by running the file through the tool with no filtering options selected.

      Let me know if you need any further help with this

      Cheers,
      Matt

  • gadicherla ramya

    Hi
    This is Ramya. I am new to using this tool and i am happy the way it works and delivers the results.
    I have a small query regarding the mirCAT tool. When i went through the data of i have some queries:

    i would like to have information on the fields like difference between the sequence and miRNA*.
    Also i have noticed that few of the miRNA[novel] are having the same sequence and also the same hairpin sequence but coding at different positions. Can you please explain this?Also some of the miRNA sequences are coded by two different chromosomes? what does this mean?
    Please let me know.

    Thanks in advance,
    Ramya.

  • The sRNA Workbench

    Hi Dimitar,

    Yes that one is ghostscript. I would prefer to have 64bit binaries for all the dependent programs for the workbench and allow it only to run in 64bit mode. Unfortunately the issue was related to packaging the static dependencies that certain programs we use rely on.

    Essentially, I always wanted the software to be a download and use (as much as possible) without the user having to install other dependencies. Unfortunately to do this there needed to be static links to certain C libraries, some of which only have a 32bit version available on Linux. Therefore, in order to build the standalone version there needed to be some 32bit programs included.

    I understand the frustration (believe me, if there is/was another way I would welcome it as it would make my life much easier!) It still means that linux users will have to install 32bit libraries on their machines (windows and OSX come with these anyway) and often I will get crash reports from those that haven’t.

    Hopefully, there will be a better solution soon as I will be removing the need for some of these external programs in a set of planned overhauls to the software.

    I hope this helps, let me know if you need any further info,

    Best wishes,
    Matt

  • Megha

    Can we use this workbench for whole Genome paired end data ( Fastq file)

    • The sRNA Workbench

      Hi Megha,

      Yes all analysis tools can be used with paired end data. However, some modifications will be needed to the adapter removal tool before FASTQ data can be processed. If you trim your files with another program to produce FASTA data, all further steps will remain the same.

      I will look into adding paired end adapter removal in a future update!

      Cheers,
      Matt

A suite of tools for analysing micro RNA and other small RNA data from High-Throughput Sequencing devices