Help Running ABACUS

[David Gemperline]

Hello,

I have downloaded and tried to run ABACUS version 2012-05-08 from sourceforge to compare two samples run through the TPP. The samples were searched with xtandem, converted to pepxml with peptide prophet, and converted to protxml with protein prophet. The combined file was made by combining the pepxml files using iprophet and then running the combined pepxml file through protein prophet. I have no problem running ABACUS with the sample dataset provided, so I assume it may have something to do with my processing in the TPP pipeline. I am not sure exactly on how to create the combined prot file. If I select the combined file as one of my uncombined protxml files, it run past the 90% mark and stalls. I think my data is formatted correctly other than the combined prot xml file. Any help would be greatly appreciated.

when i run java-d64 -verbose -jar abacus.jar

it gets stuck on mapping peptides to proteins (combined), errors out and is not consuming any CPU resources.

I get the error:

[Loaded java.util.Formattable from C:\Program Files\Java\jre7\lib\rt.jar]

Jun 12, 2012 1:06:13 PM abacus.abacusUI.workThread run

SEVERE: null

java.sql.SQLException: statement is not in batch mode

        at org.hsqldb.jdbc.Util.sqlException(Unknown Source)

        at org.hsqldb.jdbc.Util.sqlException(Unknown Source)

        at org.hsqldb.jdbc.Util.sqlExceptionSQL(Unknown Source)

        at org.hsqldb.jdbc.JDBCPreparedStatement.executeBatch(Unknown Source)

        at abacus.hyperSQLObject.makeTempProt2PepTable(hyperSQLObject.java:867)

        at abacus.abacusUI.workThread.run(abacusUI.java:2831)

Caused by: org.hsqldb.HsqlException: statement is not in batch mode

        at org.hsqldb.error.Error.error(Unknown Source)

        at org.hsqldb.error.Error.error(Unknown Source)

        ... 4 more

[Loaded org.hsqldb.lib.LongKeyHashMap$Values from file:/C:/Spcnt/abacus.jar]

[Loaded org.hsqldb.lib.FrameworkLogger from file:/C:/Spcnt/abacus.jar]

[Loaded org.hsqldb.lib.BasicTextJdkLogFormatter from file:/C:/Spcnt/abacus.jar]

A previous user suggested utilizing the clear button to fix a similar problem; however, this has not resolved my issue.

Any ideas?

[dfermin]

Hi David,

Sorry to hear you are having issues with ABACUS.

One problem I see is that you are mixing iProphet results with regular prophet results. You shouldn't "mix-and-match" result types for ABACUS.

Try this:
To make the combined file for ABACUS use just ProteinProphet.
If you are using the command prompt execute a command similar to this from within the folder that contains your 2 pepXMLs from the independent experiments:

ProteinProphet.exe *.pep.xml interact-COMBINED.prot.xml

Use the COMBINED file produced by this process for ABACUS.



A second note: You need one pepXML file for each independent experiment. A pepXML file for the COMBINED protXML will screw up Abacus.

So if you are doing 2 comparisons the files you need (in this hypothetical example) are:
interact-expt1.pep.xml
interact-expt1.prot.xml
interact-expt2.pep.xml
interact-expt2.prot.xml
interact-COMBINED.prot.xml
your_sequence_database.fasta


Let me know if any of this resolves your issues.

Damian

[David Gemperline]

Thank you!

That did the trick. It may be useful for others using Abacus who are less familiar with with the TPP and peptide/protein prophet to provide a section on how to generate the combined file.

Thank you so much, I tried troubleshooting this for days and you fixed it in less than 2 minutes.

[dfermin]

Yes a tutorial would be good.
I can tell you a small JAVA app is in the works that will facilitate the creation of the combined file.
You would select the Path to your various TPP experiments and provide unique identifiers for them.
The program would then generate the necessary interact-COMBINED.prot.xml file for ABACUS.
The main sticking point at this juncture is my lack of experience with the Windows operating system.

I would post the software on the ABACUS website as soon as it works.
In the mean time, feel free to posts any and all of your questions here. I do my best to respond to people within 24 hours.

Damian



On Tuesday, June 12, 2012 2:30:15 PM UTC-4, David Gemperline wrote:

[Brett Phinney]

Hi, just wondering if this java app to make the combined file is available now

Cheers

Brett

[Damian]

The java app to combined files is still a work in progress.
At this juncture the program can group together the necessary pepXML files into a directory and that's about it. Java is a headache when you want to launch separate applications from within it.

Someone on the TPP forum (https://groups.google.com/forum/?fromgroups#!forum/spctools-discuss)
can probably tell you a way to combine results through the GUI interface.

Through the command line you have to amass your desired pep.xml files into one folder and run:

ProteinProphet *.pep.xml interact-COMBINED.prot.xml

From within that folder. (On Windows the executable is ProteinProphet.exe)

After the program finishes you should see several new files, one of which is interact-COMBINED.prot.xml.

I hope this helps.

-- 
Damian Fermin, Ph.D.
damian...@gmail.com
http://damianfermin.com

[Brett Phinney]

Thanks, I used the TPP to make a combined file and abacus seemed to run fine and generated an output file without an error. One thing that didn't quite work is that I had 3 separate pep.xml and prot.xml files in the directory but the abacus output did not list them separately in the output file.Is there a particular naming convention abacus is looking for?

Cheers

Brett

[Damian]

I've tried to anticipate many naming conventions in the code but I can't catch them all.

Abacus prefers files to be named in this syntax: interact-your_label_here.pep.xml and interact-your_lable_here.prot.xml
This is the standard naming convention used by the TPP when specific names are given.

The program assumes that for every pep.xml file, there is a corresponding prot.xml file with the same name.
Before you run Abacus, you should check to make sure that you have 1 more prot.xml file than you do pep.xml files. The extra prot.xml file being the interact-COMBINED.prot.xml or whatever you called the merged file.
The interact-COMBINED.prot.xml file should *NOT* have a corresponding pep.xml file for Abacus to consider. Such a pep.xml file will cause serious problems with Abacus.
The best thing to do is delete the interact-COMBINED.pep.xml file if it exists.

What are your file names like?

Damian



---------------------------------------------------------------

Damian Fermin, Ph.D.
damian...@gmail.com
http://damianfermin.com

[Brett Phinney]

Thanks, so I had them named

interact.yourlabelhere.pep.xml

interact.yourlablehere.prot.xml

I'll give it a shot with an underscore

Cheers

Brett

[Damian]

The underscores are completely optional.
Email me a list of what your files are named so I can take a look.

---------------------------------------------------------------
Damian Fermin, Ph.D.
damian...@gmail.com
http://damianfermin.com

[Erik Malm]

Hi, I already have a Java app set up to interact with ProteinProphet in our in-house pipeline software, I'd be happy to create a slimmed down version to acommodate the needs for abacus. Could you just give me a short list of requirements and I might be able to get it to you in a day or so...

[dfermin]

Hi.
Thanks for the offer! This would be great!
To run Abacus you need to combine the all of the PeptideProphet pep.xml files for the individual experiments you want to analyze.
For example: if you had 3 experiments you wanted to process with Abacus you'd do:

ProteinProphet interact-expt1.pep.xml interact-expt2.pep.xml interact-expt3.pep.xml interact-COMBINED.prot.xml

This would produce the combined ProteinProphet prot.xml file from the command line.

That's it.  Let me know if you have any questions or if this isn't enough to go on.
Damian