Picard tools MergeBamAlignment error

Question

[ea11g10]

Hi all,

I am attempting to go through the Dropseq pipeline, but I have changed a few things to the default. I have aligned my fastq files to the hg38 genome rather than hg19 and I've also used TopHat to align rather than STAR. However, when I get the the MergeBamAlignment step, used to merge an unaligned bam and the aligned bam to re-introduce the tags into the aligned bam files, I keep getting an error but unsure how to resolve it.

Both the bam files are sorted by queryname as the pipeline says to do, but I keep getting the following error (I've removed some of the chromosome names otherwise it would have been too long, as it contains all the contigs):

    Exception in thread "main" java.lang.IllegalArgumentException: Do not use this function to merge dictionaries with different sequences in them. Sequences must be in the same order as well. Found [1, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 2, 20, 21, 22, 3, 4, 5, 6, 7, 8, 9, ...].
 at htsjdk.samtools.SAMSequenceDictionary.mergeDictionaries(SAMSequenceDictionary.java:305)
 at picard.sam.SamAlignmentMerger.getDictionaryForMergedBam(SamAlignmentMerger.java:197)
 at picard.sam.AbstractAlignmentMerger.mergeAlignment(AbstractAlignmentMerger.java:346)
 at picard.sam.SamAlignmentMerger.mergeAlignment(SamAlignmentMerger.java:181)
 at picard.sam.MergeBamAlignment.doWork(MergeBamAlignment.java:282)
 at picard.cmdline.CommandLineProgram.instanceMain(CommandLineProgram.java:205)
 at picard.cmdline.PicardCommandLine.instanceMain(PicardCommandLine.java:94)

The picard code I'm using is:

picard MergeBamAlignment UNMAPPED_BAM=4571121.blue101/temp/unaligned_mc_tagged_polyA_filtered.bam ALIGNED_BAM=4571121.blue101/temp/aligned.sorted.bam OUTPUT=4565921.blue101/temp/merged.bam REFERENCE_SEQUENCE=/scratch/ea11g10/Dropseq/hg38.fasta PAIRED_RUN=false INCLUDE_SECONDARY_ALIGNMENTS=false    CLIP_ADAPTERS=true IS_BISULFITE_SEQUENCE=false ALIGNED_READS_ONLY=false MAX_INSERTIONS_OR_DELETIONS=1 READ1_TRIM=0 READ2_TRIM=0 ALIGNER_PROPER_PAIR_FLAGS=false SORT_ORDER=coordinate PRIMARY_ALIGNMENT_STRATEGY=BestMapq CLIP_OVERLAPPING_READS=true ADD_MATE_CIGAR=true VERBOSITY=INFO QUIET=false VALIDATION_STRINGENCY=STRICT COMPRESSION_LE
VEL=5 MAX_RECORDS_IN_RAM=500000 CREATE_INDEX=false CREATE_MD5_FILE=false GA4GH_CLIENT_SECRETS=client_secrets.json

I created the dict file for the hg38.fasta file using picard CreateSequenceDictionary

Any help would be appreciated. Thanks

[Pierre Lindenbaum]

Both the bam files are sorted by queryname

I think, only the UNMAPPED_BAM should be sorted on queryname

[ea11g10]

In the Dropseq pipeline, after alignment of the fastq file, it says to sort the outputted alinged BAM by queryname before it moves onto the merging of the 2 BAM files