SEQanswers

Go Back   SEQanswers > Bioinformatics > Bioinformatics



Similar Threads
Thread Thread Starter Forum Replies Last Post
Trimmomatic Paired End - Low number of surviving reads BADE Bioinformatics 17 10-29-2014 09:37 AM
Illumina Paired End reads are complementary mchandra Illumina/Solexa 3 07-21-2014 09:57 AM
Illumina paired-end reads... naragam General 3 06-28-2012 05:51 AM
Illumina Reads Quality Control and Preprocessing Lisa0508 Bioinformatics 0 12-08-2011 12:06 AM
Limiting Illumina Paired-End Reads cryptic_star Bioinformatics 1 06-21-2010 05:30 AM

Reply
 
Thread Tools
Old 05-15-2015, 08:05 AM   #1
nmerienn
Member
 
Location: Switzerland

Join Date: Sep 2014
Posts: 12
Default Paired-end Illumina reads preprocessing with Trimmomatic

Dear all,

I have few questions about the trimming of adapters with Trimmomatic. I have 2x150 bp paired end runs from Illumina MiSeq system. Normally, based on FastQC, there should not be a lot of remaining adapters on the sequencing results (not detected as overrepresented sequences) . However, I think it should be better to use Trimmomatic to trimm the remaining adapters. But I am not sure to really understand how it works. If I understand well, for paired-end reads, it will simulate an adapter at the end of each forward and reverse reads based on the fasta file supplied, and then hybridize the two reads by their common features (i.e, the nucleotides corresponding to the sequenced gene). If the adapter of one read have alignment on the other one, it will be trimmed on the read. Is it true?
My problem is for the keepBothReads option. I don't understand when I should use it. At the end of the trimming and quality filtering, I would like to get two files, one for forward and one for reverse, to do the alignment with Bowtie 2. In addition, I tried Trimmomatic with and without keepBothReads option and this gave me different results: I had trimmed reads for the forward sequences without keepBothReads (around 16%), however, there was no more trimmed reads if I used the keepBothReads option. Could anyone just explain me when I should use or not this option?

Here are the command lines I used:

Without keepBothReads:
java -jar ../../../../../soft/Trimmomatic-0.33/trimmomatic-0.33.jar PE 110413_6_JY747_GCCAAT_L001_R1_001.fastq 110413_6_JY747_GCCAAT_L001_R2_001.fastq forward_true_paired.fastq forward_true_unpaired.fastq reverse_true_paired.fastq reverse_true_unpaired.fastq ILLUMINACLIP:../../../../../soft/Trimmomatic-0.33/adapters/TruSeq3-PE.fa:2:30:10

With keepBothReads:
java -jar ../../../../../soft/Trimmomatic-0.33/trimmomatic-0.33.jar PE 110413_6_JY747_GCCAAT_L001_R1_001.fastq 110413_6_JY747_GCCAAT_L001_R2_001.fastq forward_true_paired.fastq forward_true_unpaired.fastq reverse_true_paired.fastq reverse_true_unpaired.fastq ILLUMINACLIP:../../../../../soft/Trimmomatic-0.33/adapters/TruSeq3-PE.fa:2:30:10:8:TRUE

Thank you very much for your help.

Best,
Nicolas
nmerienn is offline   Reply With Quote
Old 05-15-2015, 11:12 AM   #2
dpryan
Devon Ryan
 
Location: Freiburg, Germany

Join Date: Jul 2011
Posts: 3,480
Default

You can just always use the keepBothReads options. If the size of a sequenced fragment is such that there's read through on both reads, then one read is completely redundant. By default, one read is simply tossed, which the keepBothReads option says to keep it. The benefit to keeping it is that you can still align things as paired-end without needing to do an additional unpaired alignment followed by a file merge. Bowtie2 doesn't care if reads overlap completely like this, so it won't give you any problems. The only reason to not use this option is if you're using a rather old tool that can't handle completely overlapping mates.
dpryan is offline   Reply With Quote
Old 05-15-2015, 09:09 PM   #3
nmerienn
Member
 
Location: Switzerland

Join Date: Sep 2014
Posts: 12
Default

Hi dpryan,

Thank you for your answer. So, if I understand well, if both reads are crossing the full sequenced fragment length, then the reverse has no more informations and I can discard it. But in this case, are the two output paired forward and reverse files still paired? However, in case where the overlap between the two reads is not complete, it is better to keepBothOptions to still have paired forward and reverse reads, true?
So, is it normal that the number of trimmed sequences is different between both conditions?
Thank you very much for your help.

Nico
nmerienn is offline   Reply With Quote
Old 05-18-2015, 03:30 AM   #4
usad
Member
 
Location: aachen

Join Date: Sep 2009
Posts: 53
Default

Hi Nicolas

the keeppaired is more a comfort option, so you still have paired reads. You will not loose reads by not using it, but they might end up in a different "bin" (i.e.file).

e.g.
Case1)
AAAAA-----------------------AAAAAAA
read1------------------------->

if read1 extends into the OTHER adapter then read2 can't cover other information. It can be discarded but read1 would end up in the unpaired bin which might not be the best for downstream analysis, depending on your pipeline. You can force to keep read 2 by using keepboth.
In any case forward_true_paired.fastq reverse_true_paired.fastq will always be a valid pairs.


Case2)
AAAAA-----------------------AAAAAAA
read1-------------------->

if read1 doen NOT extend into the other adapter then BOTH read1 and read2 are kept regardless of keepboth as there is new information. (unless of course read2 fails for other reasons e.g. being NNNNNNNNNN only, extremely unlikely but just in case).

Best
Bjr÷n
usad is offline   Reply With Quote
Old 05-20-2015, 12:24 AM   #5
nmerienn
Member
 
Location: Switzerland

Join Date: Sep 2014
Posts: 12
Default

Hi Bjr÷n,

Thank you for your answer, it is more clear for me now. So I kept both reads for my analysis and it seems to work well.

Best,
Nico
nmerienn is offline   Reply With Quote
Old 06-10-2015, 01:24 AM   #6
qadrikazmi
Junior Member
 
Location: Pakistan

Join Date: Dec 2014
Posts: 1
Default Trimmomatic Quality Trimming

Different Leading and Trailing Values Results in No Trimming at all...Does anybody have an idea why?
qadrikazmi is offline   Reply With Quote
Reply

Tags
illumina, paired-end, preprocessing, trimmomatic

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off




All times are GMT -8. The time now is 09:08 AM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2019, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO