OutOfMemoryError: GC overhead limit exceeded

User Questions
1

Dear All,
I am very new to MEGAN. I have metagenomics shotgun data. After quality filtering, I started with DIAMOND workflow.
I downloaded env_nr database and run makedb in DIAMOND followed by :
diamond blastx -d nr -q GRC17.fa -o --da GRC17.daa
Then in MEGAN I executed the following command:
./daa2rma -i ~/Desktop/data/grc/diamond-master/GRC17.daa -o GRC17.rma -g2t gi2tax.bin
I got the following error message
Version MEGAN Community Edition (version 6.13.5, built 7 Jan 2019)
Copyright © 2018 Daniel H. Huson. This program comes with ABSOLUTELY NO WARRANTY.
Loading ncbi.map: 2,031,704
Loading ncbi.tre: 2,031,708
Opening file: gi2tax.bin
In DAA file: /home/grc/Desktop/data/grc/diamond-master/GRC17.daa
Output file: GRC17.rma
Classifications: Taxonomy
Parsing file: /home/grc/Desktop/data/grc/diamond-master/GRC17.daa
10% 20% 30% 40% 50% 60% 70% 80% 90% 100% (4623.0s)
Total reads: 4,746,551
Alignments: 37,768,334
100% (0.0s)
Binning reads: Initializing…
Initializing binning…
Using ‘Naive LCA’ algorithm for binning: Taxonomy
Binning reads…
Binning reads: Analyzing alignments
Exception in thread “main” java.lang.OutOfMemoryError: GC overhead limit exceeded
at java.lang.String.substring(String.java:1969)
at megan.parsers.sam.Diff.decode(Diff.java:125)
at megan.parsers.sam.Diff.getReference(Diff.java:46)
at megan.parsers.sam.SAMMatch.computeAlignmentPair(SAMMatch.java:978)
at megan.parsers.sam.SAMMatch.computeAlignment(SAMMatch.java:860)
at megan.parsers.sam.SAMMatch.getBlastXAlignment(SAMMatch.java:576)
at megan.parsers.sam.SAMMatch.getBlastAlignmentText(SAMMatch.java:356)
at megan.rma6.MatchBlockRMA6.getText(MatchBlockRMA6.java:227)
at megan.rma6.MatchBlockRMA6.getPercentIdentity(MatchBlockRMA6.java:152)
at megan.algorithms.ActiveMatches.compute(ActiveMatches.java:50)
at megan.algorithms.DataProcessor.apply(DataProcessor.java:235)
at megan.core.Document.processReadHits(Document.java:538)
at megan.rma6.RMA6FromBlastCreator.parseFiles(RMA6FromBlastCreator.java:298)
at megan.tools.DAA2RMA6.createRMA6FileFromDAA(DAA2RMA6.java:340)
at megan.tools.DAA2RMA6.run(DAA2RMA6.java:303)
at megan.tools.DAA2RMA6.main(DAA2RMA6.java:64)

When I checked the other posts, It represents the memory error. My RAM is 128 GB
How do I edit MEGAN.vmoptions file? I have attached my MEGAN.vmoptions fil. Kindly help meMEGAN.vmoptions (253 Bytes)

2

Hi,

the line with -Xmx2000M sets the maximum amount of RAM MEGAN is allowed to use in your .vmoptions file. Please increase it. M is for MB, and G is for GB. e.g. -Xmx50G will allow MEGAN to use 50 GB of RAM.

3

Hi
Thank you for your reply. I got the rma file. When I import the rma file in MEGAN6, I did not any taxonomic classification. Where I have gone wrong?
I have attached my rma file for your reference

4

Hi,

The webserver probably rejected the rma file because of its size. Can you please rather post the command you used to produce the rma file?

Thanks.

5

I followed the following steps:

  1. Mapped with human genome to remove contamination using bowtie
  2. Retrived the unmapped paired-end reads and merged as single file
  3. In diamond for makedb, I used env_nr database (ftp://ftp.ncbi.nlm.nih.gov/blast/db/FASTA/env_nr.gz)
  4. diamond blastx -d env_nr -q GRC17.fa -o --da GRC17.daa
  5. daa2rma -i ~/Desktop/data/grc/diamond-master/GRC17.daa -o GRC17.rma -g2t gi2tax.bin
6

The problem is in the daa2rma command. You need to -a2t instead of -g2t as nr uses accessions instead of gi numbers. And you need the accession mapping file, available in http://ab.inf.uni-tuebingen.de/data/software/megan6/download/prot_acc2tax-Nov2018X1.abin.zip instead of gi2tax.bin

You can also use daa-meganizer instead of daa2rma, which will append classifications to the existing daa file, instead of creating an rma file.

Best,
Caner