Paired End Mapping using Bowtie

[adarshjose]

I am trying to map Illumina GA II E paired end reads to a reference genome.
More than 80 % of the reads map when using single end mapping, but virtually none of the reads map when I am using paired end mapping.
Can some one please have a look at the commands I am using and a typical example of a read I have pasted below ? Any suggestions are welcome.

Thanks in advance

Adarsh Jose
BCB, Iowa State University

I am using the following commands:

# paired end
bowtie Ref/ZmB73_RefGenome -1 sample_s4_1.fastq -2 sample_s4_2.fastq s4PairedMapping.map -n 2 -e 250 -l 30 -I 400 -X 460 -m 10 --time --un Unmappedpaireds4 --max Multipaireds4 -S -p 4 --verbose

# single end
bowtie Ref/ZmB73_RefGenome sample_s4_1.fastq s4SingleMapping_1.map -e 250 -l 30 -m 10 --time --un Unmappedsingles4_1 --max Multisingles4_1 -p 4
bowtie Ref/ZmB73_RefGenome sample_s4_2.fastq s4SingleMapping_2.map -e 250 -l 30 -m 10 --time --un Unmappedsingles4_2 --max Multisingles4_2 -p 4

# An example mapping of a read pair

# Single end Pair1
ILLUMINA-4750D6_00031:4:1:10005:2856#0/1 + chromosome:AGPv2:2:1:237068873:1 32492480 CGGCGATATAAAATAAACAACTCAGAAGCAAATCGACCCCAGAAGCCCGTCTTAC ffafff]_dfac\\\cccfcccfcf]d^]b`b[`R`dbb`bcccad`J]`b^R]^ 0 2:A>G

# Single end Pair2
ILLUMINA-4750D6_00031:4:1:10005:2856#0/2 - chromosome:AGPv2:2:1:237068873:1 32492560 CCATGGCATCCGCTTCCTCCTCCCAGCAGAGAAGATTAGCCAGCTTGTAGTCCTTCTTCATAGGAGG ^]dbddd[Z^^[Z\U`^Icf[fehhhaefgh_ffdffd]_ffechehfccfa`X]^]T`]a\aaaad 0

#Paired End
ILLUMINA-4750D6_00031:4:1:10005:2856#0 141 * 0 0 * * 0 0 CCTCCTATGAAGAAGGACTACAAGCTGGCTAATCTTCTCTGCTGGGAGGAGGAAGCGGATGCCATGG daaaa\a]`T]^]X`afccfhehceff_]dffdff_hgfeahhhef[fcI^`U\Z[^^Z[dddbd]^ XM:i:0

ILLUMINA-4750D6_00031:4:1:10005:2856#0 77 * 0 0 * * 0 0 CGGCGATATAAAATAAACAACTCAGAAGCAAATCGACCCCAGAAGCCCGTCTTAC ffafff]_dfac\\\cccfcccfcf]d^]b`b[`R`dbb`bcccad`J]`b^R]^ XM:i:0

[lakshmaa]

I am having the same problem. Did anyone solve it ?

Thanks

[adarshjose]

It is a parsing problem. The reads in both the files have to be in the same order. That is, if the 5th read in pair_1.fastq is A, then the corresponding 5th read in pair_2.fastq must also be A.

To illustrate:

This will work:

pair_1.fastq

1. >readA /1
2. >readB /1
3. >readC /1
4. >readD /1

pair_2.fastq

1. >readA /2
2. >readB /2
3. >readC /2
4. >readD /2

but here only the 1st read pairs- readA will get mapped.


pair_1.fastq

1. >readA /1
2. >readC /1
3. >readD /1

pair_2.fastq

1. >readA /2
2. >readB /2
3. >readC /2
4. >readD /2

[lakshmaa]

Thank you ! I will look into it

[nilmot13]

Hi,

I'm new to using Bowtie for paired-end mapping and too have encountered the same issue.

Did you manage to solve the parsing problem? or is there a way to program bowtie to come around the problem?

Many thanks

[adarshjose]

I don't think Bowtie takes care of this. Both the files must have the reads in the same order. Please have a look at my previous post.

[nilmot13]

Hmm do you have any suggestion as to how to combat this issue?

After clipping adaptor sequencing and trimming, often read2 becomes very short and poor in quality, which will get discarded. Do you think maintaining as much reads as possible would reduce this type of error?

[adarshjose]

Just split the reads after filtering into those with and without pairs. Use paired end mapping for the cases where both reads are retained after filtering and unpaired mapping for the remaining reads.

[nilmot13]

Thank you for the advise.

Sorry for replying very simple question. I'm really an experimental biologist, not an experience bioinformatician. Is there a FASTQ manipulation tool or command that allows you to apply the filter for pairs?