samtools mpipeup error: different line length in sequence 'chr1'.

[mihuzx]

hi,
I am new about samtools and bioinformatics, these days I try to call an SNP using bwa and samtools.everything went well, but at the last step, there is an error:
[fai_load] build FASTA index.
[fai_build_core] different line length in sequence 'chr1'.
[afs] 0:0.000
and my command is : samtools mpileup -uf ref.fa input.rmdup.bam |bcftools view -vcg ->test.raw.vcf
This is probably a dumb question, but what could be causing this?

thanks

[saran.ela123]

hi mihuzx ,
Did you sort & index the bam file before going to mpileup by using

samtools sort aln.bam aln.sorted
samtools index aln.sorted.bam

[dpryan]

hi,
I am new about samtools and bioinformatics, these days I try to call an SNP using bwa and samtools.everything went well, but at the last step, there is an error:
[fai_load] build FASTA index.
[fai_build_core] different line length in sequence 'chr1'.
[afs] 0:0.000
and my command is : samtools mpileup -uf ref.fa input.rmdup.bam |bcftools view -vcg ->test.raw.vcf
This is probably a dumb question, but what could be causing this?

thanks

The "different line length" error will generally mean that some of the lines in your reference sequence are longer/shorter than others. It's OK for the last line of a chromosome/contig to be shorter, but the others need to be the same length. If you're curious why this is the case, it's result of how indexing and random seeking of fasta files by samtools works (basically, the index gives the length of the chromosome, the number of nucleotides in each line and a chromosome offset, which is sufficient to calculate a offset position in a file for seeking). The solution, then is to just fix the fasta file. The "NormalizeFasta" command from Picard tools should be able to due this.

[mihuzx]

The "different line length" error will generally mean that some of the lines in your reference sequence are longer/shorter than others. It's OK for the last line of a chromosome/contig to be shorter, but the others need to be the same length. If you're curious why this is the case, it's result of how indexing and random seeking of fasta files by samtools works (basically, the index gives the length of the chromosome, the number of nucleotides in each line and a chromosome offset, which is sufficient to calculate a offset position in a file for seeking). The solution, then is to just fix the fasta file. The "NormalizeFasta" command from Picard tools should be able to due this.

thanks for your advice. this is the exact solution to my problem.
thanks to everyone.