No @hd header returned in sam file when running bwa mem

Question

[MboiTui]

Hello,

I produced sam files with the below command:

bwa mem -M -t 10\
    IndexedReference\
    ${sample}_R1.fastq.gz ${sample}_R2.fastq.gz\
    2> ${sample}_bwa.err > ${sample}.sam`

The resulting sam file doesn't have an @hd header. Example output of samtools view:

samtools view -H ${sample}.sam | head -n 5
@SQ SN:scaffold_1   LN:344421834
@SQ SN:scaffold_2   LN:300751346
@SQ SN:scaffold_3   LN:273662429
@SQ SN:scaffold_4   LN:261171726
@SQ SN:scaffold_5   LN:224480160

samtools view -H ${sample}.sam | tail -n 3
@SQ SN:Mitogenome   LN:18666
@PG ID:bwa  PN:bwa  VN:0.7.17-r1188 CL:bwa mem -M -t 10 4_align2genome/4_0_assembly/LitHap1 2_fastp/Lit01_L001_R1_trimmed.fastq.gz 2_fastp/Lit01_L001_R2_trimmed.fastq.gz
@PG ID:samtools PN:samtools PP:bwa  VN:1.16.1   CL:samtools view -H 4_align2genome/4_1_aligned/Lit01_L001.sam

As you can see there is no @HD header in the first line. This causes issues with ANGSD:

-> Inputtype is BAM/CRAM
-> We require a proper header starting with @HD for ANGSD
-> We observed: '@SQ    SN:sca' will exit

How can I get bwa mem to output the @HD file? I'd prefer this than having to add it manually.

bwa version 0.7.17
samtools version 1.16.1
angsd version 0.940

Cheers,
L

Answer 1

[GokalpC]

Your sam file is not coordinate sorted or converted to a bam. @HD tags come after a secondary tool such as samtools view or sort.

ANGSD page specifically mentions that bam files must be coordinate sorted.

These are required do be sorted according to reference.

Answer 2

[Pierre Lindenbaum]

the HD header is not generated by bwa, but you can add it using awk:

bwa (...) |  awk '(NR==1) {printf("@HD\tVN:1.6\tSO:unknown\n");}{print}' 

Theres's a PR about this: https://github.com/lh3/bwa/pull/336 , but sadly, all the PR have been ignored for years.