Tadpole error with paired end files

Question

[h.l.wong]

Hi all,

I am using Tadpole for error correction before assembly.

I was using the command line: tadpole.sh mode=correct in1=read1.fq in2=read2.fq out1=corrected1.fq out2=corrected2.fq according to Brian's suggestion here: http://seqanswers.com/forums/archive/index.php/t-61445.html

But it gave me error messages and there's no parameter as out1.

Here's the error message: java -ea -Xmx10000g -Xms1g -cp /share/apps/bbmap/35.82/bin/current/ assemble.Tadpole -Xmx10000g -Xms1g mode=correct in1=cleanS1_R1.fastq in2=cleanS1_R2.fastq out1=eccS1_R1.fastq out2=eccS1_R2.fastq k=50 Picked up _JAVA_OPTIONS: -Xmx1g Executing assemble.Tadpole2 [-Xmx10000g, -Xms1g, mode=correct, in1=cleanS1_R1.fastq, in2=cleanS1_R2.fastq, out1=eccS1_R1.fastq, out2=eccS1_R2.fastq, k=50]

Exception in thread "main" java.lang.RuntimeException: Unknown parameter out1=eccS1_R1.fastq at assemble.Tadpole.<init>(Tadpole.java:358) at assemble.Tadpole2.<init>(Tadpole2.java:67) at assemble.Tadpole.makeTadpole(Tadpole.java:78) at assemble.Tadpole.main(Tadpole.java:65)

Should I use tadpole's error correction one by one on the forward and reverse file like: tadpole.sh mode=correct in1=cleanS1_R1.fastq out=eccS1_R1.fastq k=50 tadpole.sh mode=correct in1=cleanS1_R2.fastq out=eccS1_R2.fastq k=50

However Brian said that the result would be inferior in the post at the link above. Any help for troubleshooting this error would be appreciated. Thank you.

Cheers

Alan

Answer 1

[jean.elbers]

At least with tadpole 37.65, I have used

tadpole.sh mode=correct in=read1.fq in2=read2.fq out=ecc-read1.fq out2=ecc-read2.fq

so maybe in later versions of tadpole, in1= has been replaced with in= and out1= has been replaced with out=

Also, you are using -Xmx10000g. Do you really have 10,000GB of RAM? or did you mean -Xmx10000m

[h.l.wong]

Thank you.

I tried your command line and it works but it runs out of memory. This time I tried -Xmx=110g as I have 120gb of RAM.

One thing I don't understand is that there's >100000mb available but it only uses 430mb, then claiming it runs out of memory. Here's the error log:

**Memory: max=104988m, free=81890m, used=430m**

Estimated kmer capacity:    2242069657
java.lang.OutOfMemoryError: Java heap space
    at ukmer.AbstractKmerTableU.allocInt1D(AbstractKmerTableU.java:222)
    at ukmer.HashArrayU1D.resize(HashArrayU1D.java:191)
    at ukmer.HashArrayU1D.incrementAndReturnNumCreated(HashArrayU1D.java:91)
    at ukmer.HashBufferU.dumpBuffer_inner(HashBufferU.java:196)
    at ukmer.HashBufferU.dumpBuffer(HashBufferU.java:168)
    at ukmer.HashBufferU.incrementAndReturnNumCreated(HashBufferU.java:57)
    at ukmer.KmerTableSetU$LoadThread.addKmersToTable(KmerTableSetU.java:551)
    at ukmer.KmerTableSetU$LoadThread.run(KmerTableSetU.java:477)

This program ran out of memory.  Try increasing the -Xmx flag and setting prealloc.

Is it necessary to set prealloc?

Cheers

Alan

[jean.elbers]

Truthfully, I am not sure, but if you have the computing hours to spare, I would try prealloc=t. I didn't use it, but I had allocated -Xmx1400g for the data that I was error correcting. I think by default, all BBMap tools use 85% of the available RAM, so no removing -Xmx might work. Not sure.