NGM: Header does not match the data #50
Description
Hello,
I really need help!
I mapped paired-end short reads to a fasta file using ngm and piped the output to samtools.
I used samtools sort to generate a sorted bamfile which I want to index with samtools index.
This is the command I used:
ngm -r HK1_racon1_nanopolished_genome.fa -1 cHK1_1.fq -2 cHK1_2.fq -t64 --max-read-length 150 --affine | samtools sort -l 9 -@ 10 -O BAM -o HK1.paired.sort.bam && samtools index HK1.paired.sort.bam
I got the following error message from samtools:
[bam_sort_core] merging from 310 files and 10 in-memory blocks...
[E::hts_idx_push] Region 589897413..589897563 cannot be stored in a bai index. Try using a csi index with min_shift = 14, n_lvls >= 6
samtools index: failed to create index for "HK1.paired.sort.bam": Numerical result out of range
It seems that one or more of the reads fall outside of the header reference. The maximum length of any reference sequence is 35803611, while the read that triggered the error starts at 589897413. It looks like the longest chromosome in the SQ headers is ~36Mb while the alignments are out to ~590Mb. This would appear to indicate the header does not match the data and therefore the indexing is not working. So, the issue might be with the ngm alignment.
Do you have any tipps?