Poor alignment rate by bowtie2 for exome sequence data

[ty23991]

Hi
I am aligning the paired exome sequencing data, that originated from

Sorted raw data bam file -->
Converted to fastq -->
Trimmed for adapters and quality -->
Alignment using bowtie2.

I get only 0.03% and 0.01% reads aligned concordantly >=1 times. Is there any way I can improve the alignment ? I would appreciate comments.

The following is the full alignment summary:

50608141 reads; of these:
50608141 (100.00%) were paired; of these:
50591535 (99.97%) aligned concordantly 0 times
14032 (0.03%) aligned concordantly exactly 1 time
2574 (0.01%) aligned concordantly >1 times
----
50591535 pairs aligned concordantly 0 times; of these:
86964 (0.17%) aligned discordantly 1 time
----
50504571 pairs aligned 0 times concordantly or discordantly; of these:
101009142 mates make up the pairs; of these:
50807764 (50.30%) aligned 0 times
45905252 (45.45%) aligned exactly 1 time
4296126 (4.25%) aligned >1 times
49.80% overall alignment rate

[rcook34]

Can you show us your full bowtie2 command? For exome-seq data to hg19 I always tend to get mapping rates > 95%, so I would think there is clearly a problem here. At first glance it looks like the vast majority of your paired reads have one pair aligning, with the other failing to align for some reason.

[ty23991]

Thanks
Here is the command:
bowtie2 --phred33 -x HG19 -N 0 -I 60 -1 sbx.R1_trimmed.fastq.gz -2 sbx.R2_trimmed.fastq.gz -S sbx_hg19_x.sam -p 16

[rcook34]

Command looks good to me...

Can you share some detail about your workflow? ie not sure how you're going from a sorted bam file to fastq then trimming then (re)alignment... Is this data you found somewhere or you sequenced yourself?

The only other thing I can think of is perhaps the insert size is causing problems for the alignment (ie too large). You could try modifying that parameter, as well as attempting to the alignment in single end mode to see what kind of outputs you get.

Hopefully someone else more senior sees this post and has a more concrete idea of what is going on!

[Brian Bushnell]

Sounds like you messed up the order and pairing of the reads at some point, probably when you sorted them. There's no reason to sort an unaligned bam file.

[ty23991]

Thanks all for the suggestions.
I changed the tool used for trimming. I shifted to cutadapt - performed trimming for adapters and by base quality.

Finally i got the rates of alignment for the sample set ranging between 85-97%.

Thanks to all