ERROR:SAM to BAM using samtools

Question

[Yi Chu]

I tried to convert the SAM file generated by Bowtie(1.2.3) to BAM file with samtools(1.9), but I met the following warnings and errors:

[W::sam_hdr_sanitise] Missing trailing newline on SAM header. Possibly truncated
[E::sam_hdr_error] Missing tab at line 32: "@HISEQ:154:CB8RGANXX:1:2101:1470:1000    16    Fvb7    20495637    255    16M    *    0    0    GTTGTGCTACAGCGGN    IIIIIIIIIIIIIIII    XA:i:1    MD:Z:15A0    NM:i:1    XM:i:3" 
[main_samview] failed to write the SAM header

but when I use cat -A to check the newline in my sam file, it looks normal that a $ character appears at every line end. Here is the header region of my sam file :

I think there maybe some problems with my sam file. The commands I used are as below:

bowtie-build -f ref.fasta index 
bowtie -f -v 1 -p 8 -S -a -m 50 --best --strata index s3-7-6-hua-2_clean.fa s3-7-6-hua_clean.sam --sam-RG ID:s3-7-6-hua-2_clean
samtools view -bS s3-7-6-hua_clean.sam  >s3-7-6-hua_clean.bam

Appreciated for any help!QAQ

[Pierre Lindenbaum]

"Missing tab at line 32"

but when I use cat -A to check the newline in my sam file, it looks normal that

the message is about a tab, not about a newline.

[Pierre Lindenbaum]

did you use any tab instead of spaces in the command "bowtie -f -v 1 -p 8 -S -a -m 50 --best --strata index s3-7-6-hua-2_clean.fa s3-7-6-hua_clean.sam --sam-RG ID:s3-7-6-hua-2_clean" , it could mess the @PG line in the header.

[Yi Chu]

Thank you for your help:D! It turns out that the '@' character brings the problem~

Answer 1

[John Marshall]

Your read names @HISEQ:154:etc look like SAM headers, so your SAM file is not usable. Read names in SAM files are not allowed to start with @ characters.

So either your FASTQ file somehow has two @@ characters at the start of its header lines, or your bowtie mapping pipeline is somehow leaving the @ FASTQ header character in as part of the read names.

[John Marshall]

…or which one of those bowtie arguments is the input sequence data file anyway…

[Yi Chu]

It is the @ character in the read names that leads to the error message. I modify my fasta file and then samtools works fine. Thank you so much :D