MarkDuplicates- inadequate memory

, , , ,
1

Hello, I am attempting to use markduplicates on my data aligned and sorted by Bowtie2. Everytime I attempt to run this tool, I receive the error detailed below.
“The job terminated because it used more memory than allocated”. This is the command line code
_JAVA_OPTIONS=${_JAVA_OPTIONS:-“-Xmx2048m -Xms256m -Djava.io.tmpdir=${TMPDIR:-${_GALAXY_JOB_TMPDIR}}”} && export _JAVA_OPTIONS && ln -sf ‘/corral4/main/objects/8/6/4/dataset_864f72d3-51b1-4ab5-a694-4e8925bb7933.dat’ ‘Rep_2_1-250127-A02_fastq’ && picard MarkDuplicates --INPUT ‘Rep_2_1-250127-A02_fastq’ --OUTPUT ‘/corral4/main/jobs/064/890/64890385/outputs/dataset_8e9d140c-9b3d-4a7f-acc4-194d99bcf1e4.dat’ --METRICS_FILE ‘/corral4/main/jobs/064/890/64890385/outputs/dataset_6e50595b-8618-4d4c-88be-86e49ae876af.dat’ --COMMENT ‘using BAM with read group, 1000 for nxtsq 2k’ --REMOVE_DUPLICATES ‘false’ --ASSUME_SORTED ‘true’ --DUPLICATE_SCORING_STRATEGY SUM_OF_BASE_QUALITIES --OPTICAL_DUPLICATE_PIXEL_DISTANCE ‘1000’ --VALIDATION_STRINGENCY ‘LENIENT’ --TAGGING_POLICY All --QUIET true --VERBOSITY ERROR

How do I allocate more memory to this tool?

2

Hi @DHowell23

The memory error can be produced by any tool, and is usually a clue that something might be wrong with the input data or parameters. A recent topic with more details about what to do is here, and most of that will apply for you, too. → Troubleshooting Snippy memory allocation error

Then, if the data is Ok, you can try one more option for this specific tool. Filtering your BAM file might help the job to process. Remove any unmapped and lower mapQ alignments. Maybe filter for proper pairs and the primary hit. The downstream analysis steps probably expect filtered data anyway (example: Chip-seq would). See our tutorials for how the filtering is commonly applied and why (improves results!).

If the filtering is not enough, then you can try at a different Galaxy server to see what happens. To move data from the UseGalaxy.org server to another server, maybe UseGalaxy.eu, follow this guide. You can move a single dataset or a batch of data.

Xref

Hope this helps! :slight_smile:

1 Like