Picard ERROR:INVALID_VERSION_NUMBER of sorted BAM file

Question

[marongiu.luigi]

Dear all,

I have checked some sorted BAM files with Picard 2.18.13 and I got this error:

## HISTOGRAM    java.lang.String
Error Type                      Count
ERROR:INVALID_VERSION_NUMBER    1

I am using this version of Java on an Ubuntu 18 machine:

$ java -version
java version "10.0.2" 2018-07-17
Java(TM) SE Runtime Environment 18.3 (build 10.0.2+13)
Java HotSpot(TM) 64-Bit Server VM 18.3 (build 10.0.2+13, mixed mode)

I can see from here that the problem might be Picard rather than the BAM file. But from here. the issue seems to lie with the headers: "Does not match any of the acceptable versions". The header I got is:

$ samtools view -h file1.bam | head -n 5
@HD VN:1.6  SO:coordinate
@SQ SN:21   LN:46709983
@RG ID:C3MF6ACXX.L001   SM:501  PL:ILLUMINA LB:lib-501  PU:C3MF6ACXX.1.NoIndex
@PG ID:bwa  PN:bwa  VN:0.7.17-r1188 CL:bwa mem -t 10 -R @RG\tID:C3MF6ACXX.L001\tSM:501\tPL:ILLUMINA\tLB:lib-501\tPU:C3MF6ACXX.1.NoIndex ./ref/GRCh38-21.fa 501N-1_1.fq.gz 501N-1_2.fq.gz -o aln/501N_bwa.sam
HWI-ST1437:64:C3UM1ACXX:4:2214:15629:46172  163 21  2145    0   52S19M30S   =   2145    19  ATATACATATACAAACATTCATAACAAAATAAGGAATATTTATATAATAATTACAGTCCTCATGTTAATAACTGGTCACATGCTTATAGCAGGTATTTATA   CCCFFFFFHHHHHJJJJJJJJJJJJIJJJJJIJJJJJJJJJJJJJJIJJJGIJJIJHIJJJJJJIIIHIJJJJJJIIJIIJJJJJHHHHHGGFDFFFFFEE   NM:i:0  MD:Z:19 MC:Z:27S19M55S  AS:i:19 XS:i:19 RG:Z:C3MF6ACXX.L001

$ samtools view -h file2.bam | head -n 5
@HD VN:1.6  SO:coordinate
@SQ SN:21   LN:46709983
@RG ID:C3MF6ACXX.L001   SM:502T PL:ILLUMINA LB:lib-502  PU:C3MF6ACXX.1.NoIndex
@PG ID:bwa  PN:bwa  VN:0.7.17-r1188 CL:bwa mem -t 10 -R @RG\tID:C3MF6ACXX.L001\tSM:502T\tPL:ILLUMINA\tLB:lib-502\tPU:C3MF6ACXX.1.NoIndex ./ref/GRCh38-21.fa 502N-1_1.fq.gz 502N-1_2.fq.gz -o aln/502N_bwa.sam
HWI-1103R:166:C3UJNACXX:1:1216:10747:27634  163 21  2359    0   48S19M34S   =   2359    19  CCCCTGGGTTAATCTACTTCTCATTATAAACGATGACCCATGATGATGTGTTTTCTGAACCACTCATCTTGCATAGAGTGCCACAATGTGGAACAGCCCTA   C@CFFFFFHHGHHJIIIJJJIGIIJJJJJJJJIIGGIIJIHIGIJIJIFHGHIJJJJIJIJJIJIJGGIIJIFGHHHGEHEFEFFCCECDCECCBDDDDCD   NM:i:0  MD:Z:19 MC:Z:9S19M73S   AS:i:19 XS:i:19 RG:Z:C3MF6ACXX.L001

The problem is there even if I align with HISAT2:

$ samtools view -h fle2HST.bam | head -n 5
@HD VN:1.0  SO:coordinate
@SQ SN:21   LN:46709983
@RG ID:C3UJNACXX.L001   SM:502  PL:ILLUMINA LB:lib-502  PU:C3UJNACXX.1.NoIndex
@PG ID:hisat2   PN:hisat2   VN:2.1.0    CL:"/home/gigiux/src/hisat2/hisat2-align-s --wrapper basic-0 --rg ID:C3UJNACXX.L001 --rg SM:502 --rg PL:ILLUMINA --rg LB:lib-502 --rg PU:C3UJNACXX.1.NoIndex -p 10 -q -x ref/GRCh38-21.fa -S aln/502N_hst.sam -1 /tmp/14444.inpipe1 -2 /tmp/14444.inpipe2"
HWI-1103R:166:C3UJNACXX:1:2307:16976:65381  163 21  5010021 1   101M    =   5010148 228 CTAAAGTGCTGGGATTACAGGTGTTAGCCACCACGTCCAGCTGTTAATTTTTATTTAATAAGAATGACAGAGTGAGGGCCATCACTGTTAATGAAGCCAGT   CCCFFFDDHHHHHJJJJJJJJCFHIIJJJJJJIJJHIIJJJJJIJJIJJJJJJJJJIJJJIIJJJIJIJJJJCHHHHFFFDEEEEECDCDEEDDDDDDDD:   AS:i:0  ZS:i:0  XN:i:0  XM:i:0  XO:i:0  XG:i:0  NM:i:0  MD:Z:101    YS:i:0  YT:Z:CP RG:Z:C3UJNACXX.L001 NH:i:2

Would you know how can I solve it?

Thank you.

[Devon Ryan]

To call picard overly picky would be an understatement. Whenever possible use something else.

[ATpoint]

I could not agree more :)

[Pierre Lindenbaum]

picard: the command line arguments are very unpleasant (nevertheless you can read them from a file) but otherwise, I like it. (le's start a troll with picard :-) )

[marongiu.luigi]

I have seen a lot of complaints about Picard on the web; but what is the alternative then? I understand Picard is a required as a preparatory step to GATK. Is there something better? I have been told sambamba, but this looks to me more a replacement for samtools.

[Devon Ryan]

If you need read groups then have your aligner add them. If you need to sort or mark duplicates, that's what sambamba or samtools are for. Picard is then no longer needed.

[marongiu.luigi]

OK then, I'll try sambamba. Do I need to run an error check? I haven't seen a command similar to ValidateSamFile in sambamba.

[Devon Ryan]

You shouldn't normally use that command, it's only appropriate when you get an error and even then it's easier to samtools quickcheck, which ensures the basic file is intact and doesn't do all of the absurd stuff picard tries to do.

Answer 1

[Pierre Lindenbaum]

@HD VN:1.6 SO:coordinate

picard use the java library htsjdk, your error was thrown from https://github.com/samtools/htsjdk/blob/master/src/main/java/htsjdk/samtools/SamFileValidator.java#L563

because 1.6 is not supported by htsjdk https://github.com/samtools/htsjdk/blob/master/src/main/java/htsjdk/samtools/SAMFileHeader.java#L52

 public static final Set<String> ACCEPTABLE_VERSIONS = CollectionUtil.makeSet("1.0", "1.3", "1.4", "1.5");

a quick hack should fix the problem:

sed '/^@HD/s/1\.6/1\.5/'

[marongiu.luigi]

Thank you. The problem here is that the files are in BAM format: how do I use sed on binary files? Also, how happen this error does not appear on other files treated in the same way?

[marongiu.luigi]

I used the brute force approach: I reconverted the BAM into SAM, changed the line and then reconverted into BAM. Now no errors.

Is there a quicker approach? Is it possible to modify a BAM file directly? And why this error does not come out for all files?

Tx