PAREsnip

Available from Version: 2.3

PAREsnip is a fast user-friendly multi-platform bioinformatics tool to find targets of small RNAs using the degradome. A full PDF manual for PAREsnip is available here

System requirements

Minimum: Java SE 6 /  6 Gb RAM Recommended: Java SE 6 / 16 Gb RAM

Starting PARESnip

Choose PAREsnip from the Tools menu in the Workbench. To start the tool from the command line, navigate to the Workbench installation directory and issue the command

java -jar Workbench.jar -tool paresnip

The syntax for providing inputs to the tool will be displayed.

PARESnip window

The PARESnip application after performing an analysis. Additional information and statistics for each of the input data sets is provided in the information area (top left box). This information includes the number of sequences and sequence length distributions. In the top-right box, shorter informational and instructional messages are provided to help guide a user through the analysis. The estimated time remaining to complete processing is also shown. Small RNA/target interactions are shown in the ‘Output’ table. The table can be sorted by clicking on a column header. The column order may be changed using drag and drop. The search box at the bottom-left can be used to search the table – partial strings are matched and case is ignored.

PARESnip menus

File menu

  • Select Open to choose the input files.
  • Select Save… or Save as… to save results.
  • Select Reset to reinitialise PARESnip.
  • ‘Close’ will close the window

Run menu

  • Select Start to begin processing
  • Select ‘Cancel’ to stop execution.

View menu

A description of the columns in the result file

Gene – the ID or description of the mRNA/transcript.
Category – Potential cleavage sites on a single transcript can be categorized according to degradome read abundance. Category 0 is the strongest signal and 4 the weakest.
Cleavage Position – the position on the transcript where cleavage has been identified.
P-Value – a score that indicates how likely the reported duplex occurred by chance.
*Fragment Abundance – the abundance for the degradome fragment (tag) indicating cleavage at that position.
*Weighted Fragment Abundance – calculated by dividing the abundance of a degradome fragment (tag) by the number of positions across all transcripts to which the tag has aligned.
*Normalized Weighted Fragment Abundance – weighted Fragment abundance/total number of fragments x 1,000,000.
Duplex – the sRNA|mRNA alignment.
Alignment Score – within the duplex, a mismatch contributes 1.0 to the score, unless it is a G-U (wobble) pair in which case it contributes 0.5 to the score. A gap in the alignment contributes a value of 1.0 to the score.
Short Read ID – the ID of the sRNA.
Short Read Abundance – the abundance of the sRNA.
Normalized Short Read Abundance – sRNA abundance/total number of sRNAs x 1,000,000.

  • Kun Yang

    This problem happened to me too. 8G RAM I have and “-Xms6g -Xmx6g” was used to run workbench. While it running, I observed that only 3G RAM was used but “An out of memory error occurred in ExactMatcher” was sent to me. Swap was not used to. My OS is Ubuntu 15.04. JAVA is JDK 1.8.0_51 download from oracle.

    • The sRNA Workbench

      Hi Kun,

      Did the software run with the test data?

      Cheers,
      Matt

      • Kun Yang

        Yes, I followed your tutorial vedio and used test data in “TutorialDataV3” while my system monitor open. I observed that 3G RAM was used but told me “An out of memory error occurred in ExactMatcher”.

  • Robert King

    Thanks Matt, is there an ETA on the next version?

    • The sRNA Workbench

      Hi Robert,

      Hopefully I will be releasing an alpha of the next version in a week or so, this is a bit of a test run for some fairly sweeping changes to the entire program so I am not sure how many of the actual new features to older tools will make it in.

      I really want to just release the new tools to get them being used as soon as possible and then work on the old stuff while the new bugs get worked out!

      Watch this space…

      Cheers,
      Matt

  • Robert King

    If you don’t have degradome results can you still identify targets of small RNA using some other method?

    • The sRNA Workbench

      Hi Robert,

      Sorry for the delay in replying to this, I only just spotted it when you posted you last comment!

      Currently not in the workbench, I am investigating other target prediction tools at the moment to add and will make them available in an upcoming version.

      Cheers,
      Matt

  • Igor Yakovlev

    Hi,
    I have The UEA small RNA Workbench Version 3.2, but started from Workbench.jar (not UEA_sRNAWorkbench.exe, it does not start from application file). PARESnip tool runs out of memory immediately after start. I have 64 Gb of memory, so I do not think it is the memory problems but maybe some configuration. Could you, please, recommend something what could be done to make it work

    Thank you in advance
    Igor

    • The sRNA Workbench

      Hi Igor,

      If you start from the workbench.jar you must specify how much RAM you will give the software yourself (otherwise it just has the default amount).

      To do so you must launch the software from the command line and type the command:

      java -Xms10g -Xmx10g -jar Workbench.jar

      to give the software 10GB of RAM for example.

      Let me know if you need any further help and if this works for you!

      Cheers,
      Matt

  • The sRNA Workbench

    Hi Jaes,

    Could you let me know exactly what the error message reported to you?

    Cheers,
    Matt

  • Arash

    Hi,
    PAREsnip is run only on windows? can it be run on Ubuntu or Mac?

    • The sRNA Workbench

      Hi Arash,

      the code should run under linux and OSX. Is there an issue getting it to work?

      Cheers,
      Matt

      • Arash

        Thank you, there was no issue, in the meantime I could not find any tutorial data for PAREsnip here, there is only a PDF manual with no tutorial data, is there any file in the TutorialData that can be used as PAREsnip inputs (degradome, transcriptome and miRNA query)?

        • The sRNA Workbench

          Hi Arash,

          Yes there is some tutorial files you can use. If you have downloaded the tutorial data from source forge you can use the data found in the FASTA directory, there is one file called oneTestDeg.fa a miRNA file called oneMicroRNA.fa and a transcript called oneMRNA.fa

          Those files should allow for example execution. There is no tutorial video ready for this program yet though.

          Let me know if you need any further help with this.

          Best wishes,
          Matt

          • Arash

            Thank you very much, I could run PAREsnip on those files, but before that we have to identify the name of sequences in the TestDeg.fa (for example read 1, read 2, …), PAREsnip will be executed and gives two hits.

          • The sRNA Workbench

            Ok great, I will add this information to the tutorial when it is ready to help others (unless they see this comment!)

            Best wishes,
            Matt

  • The sRNA Workbench

    Hi Vikas,

    Sorry for the delay in response I have been away on holiday last week.

    At present there is no import option. I can add one if you think this would help?

    Cheers,
    Matthew

    • VikasGupta

      Hi,
      Thank you for response. At least for us, import option would be really useful. Right now, we are working on many pair of smallRNA/degradome libraries and importing option will certainly ease our analysis.
      Thank you once again,
      Best,
      Vikas

  • Mars William

    Hi,
    I am a gratude student and I am interested in the PAREsnip. But I can’t find the download of the PAREsnip?Could you tell how I can download the PAREsnip?
    Thank you very much
    Mars william

    • The sRNA Workbench

      Hi Mars,

      I have sent you an email. Did the link work for you?

      Best wishes,
      Matt

      • Mars William

        Hi Matt,

        I have got it! Thanks!

        Best wishes!

        Mars

      • jaes

        can i get the PAREsnip in window OS

        • The sRNA Workbench

          Hi Jaes,

          Yes the complete software package is available for Windows. If you head to the downloads page on windows it should automatically point you at the correct download, or if not let me know on here and I will copy in a link for you.

          Cheers,
          Matt

          • jaes

            There are limit input of UEA sRNA workbench especially for PAREsnip because since i run in 64Gb of Ram with 24 core it is still can’t support 151000kb of degradome and 22100kb of transcriptomic. can you give suggestion of this

          • The sRNA Workbench

            Hi Jaes,

            There are limits but 64GB should be enough for that analysis I think. How did you set up the software? Did you allow the startup program to auto detect the RAM? If so how much did you have free at the time (did you have any other programs running or was a lot allocated to disk cache) or did you configure by hand?

            Cheers,
            Matt

  • Igor Yakovlev

    Hi,
    I am working with piRNAs and tried to run PARESnip, but the upper limit for sRNAs is 24 nt. So piRNAs not suitable for the system. Is it possible to loose upper limit of sRNA until 34-36 nt so it will be possible to check putative targets of peRNAs also?! Thank you
    Igor Yakovlev

    • The sRNA Workbench

      Hi Igor,

      PAREsnip assumes that cleavage is of a transcript is mediated by Ago proteins (between nucleotide positions 10 and 11 of the guide miRNA). I am not an expert on piRNAs but have just done some reading (http://www.nature.com/nrg/journal/v14/n8/full/nrg3495.html) and it appears that after piRNAs are incorporated into a Piwi protein that the sequence not protected by the protein is trimmed back from the 3′ end yielding a shorter processed piRNA. As far as I can tell the unknown here is that I am not sure if cleavage would still occur at positions 10-11 which, if not, would make PAREsnip useless for this analysis. In terms of a length limit you should be able to get around this by pre-trimming your piRNAs (from the 3′ end) to a length within the range that PAREsnip would accept (since the cleavage position is counted from the 5′ end anyway it shouldn’t matter how long your sequence is). I’m not saying that this is sensible biologically speaking but it could be a workaround.

      Cheers,
      Matt

      • Igor Yakovlev

        Hi, Matt
        Thank you for your quick answer. I am also not sure where the cleavage occur, some author mark it at 10-11 position, some mark in context with T/A somewhere in the middle, some did not mark at all. I will try to trim from 3’ end until 24 nt.
        Vennlig hilsen/Yours sincerely
        Igor Yakovlev
        Seniorforsker/ Senior Research Scientist
        – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – Norsk institutt for skog og landskap / Norwegian Forest and
        Landscape Institute
        Hogskoleveien 8, NO-1431 Ås
        T (+47) 64 94 91 71
        F (+47) 64 94 29 80
        – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – http://www.skogoglandskap.no
        – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –

  • BMW

    Besides, after output the results, it couldn’t save but replace a file that i preset in linux, but sometimes it doesn’t work.

    Awaiting your reply.

    thanks

    miao

    • The sRNA Workbench

      Hi Miao,

      Just to let you know, I have forwarded your questions onto the tools original developer and I will let you know on here as soon as I have a response for you.

      Cheers,
      Matt

  • BMW

    HI Matt,

    I downloaded and utilized the PAREsnip for searching the targets of tomato miRNAs.

    It is exciting me by the fast speed compared to Cleaveland2!

    However, in this test, we found many of targets of sly-miRNAs (conserved) couldn’t list on the “output” plane, which targets were detected in Cleaveland2 by load the same data.

    I don’t know whether it hasn’t identified or just did not shown in Linux (REDHET)?

  • Arjun

    Hi,

    I am testing PARE snip to find targets of Arabidoopsis miRNAs available at miRBASE. I have degradome file in fasta format and using genomic trnascripts for ‘-trans file’. But the problem is that PAREsnip reports no validated target:

    Small RNA File Filter Stats:

    Read Counts: TOTAL NR

    input 338 271

    filter by sequence length 338 271

    filter low-complexity sequences 338 271

    filter by min-abundance and max-abundance 338 271

    filter invalid sequences 0 0

    filter by t/rRNA (matches out) 0 0

    ——-

    Analysis Complete.
    Time Elapsed: h: 0 m: 6 s: 17 ms: 42
    No hits found in PlotRecords

    I am using commandline version of PAREsnip on server with no dearch of memeory or cores. Also, is there anyway I can define parameters to run PAREsnip on multiple cores. I see there is an option to define parameters file but what is the format of parameters file?

    Awaiting your reply.

    Arjun

    • Hi Arjun,

      You can find examples of all of the parameter files in data/default_params

      There is a parameter to modify for PAREsnip that will control the maximum number of threads the process can spawn.

      It looks like 338 total and 271 non-redundant (NR) miRNA sequences came out of the “min-abundance and max-abundance” filter and went into the “invalid sequences” filter, but none came out. So I would suggest that no results were produced because all the miRNAs were filtered out. This would indicate that there are unknown/invalid characters within the miRNA sequences.

      If you need further help you can send me a sample of the data (matthew.stocks@uea.ac.uk) and we can look into it for you.

      Best wishes,
      Matt

  • LEPLE Jean-Charles

    HI,

    How to get the degradomeID or degradome sequence related to a duplex result?

    sincerely

    Jean-Charles

    • Hi Jean-Charles,

      Degradome fragments are exactly matched to the transcriptome and 5′-end alignment positions are recorded as well as degradome fragment abundance at that position. As only the position and fragment abundance are retained from the degradome data, there is currently no method provided by the tool to separately output degradome IDs/sequences. However, this is something we could include with our next update.

      Best wishes,

  • Yuwen Zhou

    Is it possible to generate T-plots from the command line? I tried to load the PARESnip results (command line) to VisSR (GUI) but i got an error: “An error occured while loading data from the file.”

    • Yuwen Zhou

      And could you explain me the meaning of columns in the result file? I’m confused. It seems that the “Short Read ID” represents the ID of miRNA (I’m using miRBase miRNAs as reference), instead of degradome tags (my reads). In this case which colunm should I take as the read abundance at a certain cleavage site?
      What means the “Fragment Abundance”? In my result the maxium is 2.0 . Is it normal?
      Is it possible to output more intermediate results such as mapping result etc.

      • Hi Yuwen Zhou

        You are correct in regards to the “Short Read ID”. It is indeed the identifier for the sRNA/miRNA. The “Short Read Abundance” column shows the abundance of the sRNA. For the abundance of degradome tags (Fragments), three abundance values are provided by the tool and are described below (prefixed with *).

        It is not possible for the tool to show intermediate results like you suggest, nor is it possible to generate PAREsnip t-plots via the command line. Currently, only results generated by PAREsnip in GUI mode are compatible with VisSR in GUI mode. We appreciate your feedback and the additional functionality you have highlighted will be taken into account during future updates.

        I have added a description of the column headings in the page above to help clarify the results.

        We hope this helps. Please feel free to leave further comment if you need any more help 🙂

        Best wishes,

        • Yuwen Zhou

          Thank you very much !!!
          Best wishes,

  • David Horner

    Very impressive guys … once I got over the same memory problems described above by John it works like a dream. Much much faster than Cleaveland3 and at a first glance, the results look a bit cleaner as well. Many thanks

    • Hi David,

      Apologies for the delay in reply, I have been on holiday last week.

      Thanks very much for the message 🙂 Please let us know if you have any other issues or questions regarding the Workbench. In addition please remember to subscribe to our RSS feeds to keep up to date with future improvements to our software.

      Best Wishes,
      Matt

  • Pingback: New degradome analysis tool – PAREsnip | Domaining.in blog()

  • John White Bai

    I’m getting an error while trying to deal with Degradome data (42 million lines, 35nt short reads):
    “An out of memory error occurred in ExactMatcher”
    Any ideas?

    • John White Bai

      Ubuntu 11.04, 16G RAM
      java version “1.6.0_22”

    • admin

      Apologies for the delay in reply. The university has been closed over the Easter break.

      I have sent an email to the author of the PAREsnip tool to ask for further information on this. My first thoughts are that the data is too large to fit into memory.

      I have a few questions, did you run the program from the sRNAWorkbenchStartup.jar? Or did you run the workbench directly?

      How recently had you restarted your Ubuntu machine? Ubuntu has a system in place that reserves large amounts for a disk cache and this causes problems with our Java programs that attempt to detect and allocate “free” memory in advance to the virtual machine as Linux reports cached memory as being used (but will free it on request from software).

      If you want to ensure the maximum amount of memory is allocated to the workbench you can configure this by hand.

      Type the command (for example):

      java –Xms10g –Xmx10g –jar Workbench.jar

      to allocate 10Gb of RAM to the workbench. This may help with your memory problem. In addition, when I have further information from the author of the tool I will get back to you!

  • M. Taraka Ramji

    Is this tool animal specific or even the plants are included???kindly let me know through the email….

    • admin

      In principle, the tool would work on animal data provided you can acquire a degradome (which may prove difficult). Therefore PARESnip is most likely to be used for plant miRNA target prediction.

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