Paired-end Illumina RNA-seq adapter trimming

fabrice · 06-19-2011, 04:11 PM

Does anyone can recommend a tool for Paired-end Illumina RNA-seq adapter trimming? Most of tools are for single-end sequence. I think if there a tools used for paired-end data.
The shortcoming of the single end tools:
1, The single end tools sometimes will make the paired-end data mess.
2, It does not consider paired-end information, if there is a tool for paired-end, such remove will be accurately.

Simon Anders · 06-20-2011, 01:14 AM

HTSeq has facilities useful for this.

You can write a little Python script like this:

Code:

import itertools
import HTSeq

in1 = iter( HTSeq.FastqReader( "mydata_1.fastq" ) )
in2 = iter( HTSeq.FastqReader( "mydata_2.fastq" ) )
out1 = open( "trimmed_1.fastq", "w" )
out2 = open( "trimmed_2.fastq", "w" )

for read1, read2 in itertools.izip( in1, in2 ):
   read1.trim_right_end( "ACGGTC" )
   read2.trim_left_end( "TTCGAC" )
   read1.write_to_fastq_file( out1 )
   read2.write_to_fastq_file( out2 )
      
out1.close()
out2.close()

I haven't tested the script, and you will have to adjust it to your needs, of course.

Note that you can make the trimming tolerant to read errors by either adding a second argument, giving the allowed fraction of mismatches, or using trim_right_end_with_quals, which accepts a maximum sum of mismatch qualities. See here and here for an explanation.

fabrice · 06-20-2011, 01:47 AM

Simon Anders,
Thanks for your reply.
Could you please also give an example how to trim 3' reads which have low quality? For example, set a cutoff value. Bellow this cutoff, the nucleotide will be trimmed.

Thanks.

saharnm · 07-20-2011, 06:31 PM

Take a look at "http://solexaqa.sourceforge.net/". It works on both single and paired-end reads.

vbiaudet · 11-17-2011, 06:40 AM

We have the same problem for paired-end of 100bp and sizing of 200 (fragment size) from Illumina GAIIx RNAseq. In many case we think we got an adapter in 3' that have been sequenced... we think that because the mapping is very bad in 3'. BUT we know that the Quality score is very good (Q> 30). The problem is the 3'end can contains just some bases from the adapter and not the complete sequence... when the list of adapter is known did you find a tool to solve this problem ?

Thanks, vb

ataheri · 03-28-2012, 09:22 PM

We are using Trimmomatic as it is able to do trimming and adaptor removal for paired-end reads. At first we tried fastx-toolkit but then we switched to trimmomatic as a better choice for paied-end samples.

relipmoc · 09-23-2013, 06:21 PM

You may also try skewer which is an adapter trimmer dedicated to Illumina paired-end data

saran.ela123 · 11-18-2013, 12:54 AM

May i know from where to download skewer ?

FlorDV · 01-05-2015, 08:48 AM

Hi saran.ela123

you can download it from here: http://sourceforge.net/projects/skewer/files/

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
EBARDenovo - A new RNA-seq do novo assembler for paired-end Illumina data	htchu.taiwan	RNA Sequencing	2	06-10-2013 01:13 AM
EBARDenovo - A new RNA-seq do novo assembler for paired-end Illumina data	htchu.taiwan	Illumina/Solexa	9	04-16-2013 12:08 AM
Human Illumina Paired-end RNA-Seq remove duplication.	fabrice	Bioinformatics	8	10-15-2012 10:10 PM
RNA-Seq: Single Read and Paired End mRNA-Seq Illumina Libraries from 10 Nanograms Tot	Newsbot!	Literature Watch	0	11-09-2011 03:10 AM
RNA-seq: Replicates, single-end, paired-end story	pasta	Bioinformatics	2	07-05-2011 12:51 AM

06-19-2011, 04:11 PM	#1
fabrice Member Location: paris Join Date: Oct 2009 Posts: 86	Paired-end Illumina RNA-seq adapter trimming Does anyone can recommend a tool for Paired-end Illumina RNA-seq adapter trimming? Most of tools are for single-end sequence. I think if there a tools used for paired-end data. The shortcoming of the single end tools: 1, The single end tools sometimes will make the paired-end data mess. 2, It does not consider paired-end information, if there is a tool for paired-end, such remove will be accurately.

06-20-2011, 01:14 AM	#2
Simon Anders Senior Member Location: Heidelberg, Germany Join Date: Feb 2010 Posts: 994	HTSeq has facilities useful for this. You can write a little Python script like this: Code: import itertools import HTSeq in1 = iter( HTSeq.FastqReader( "mydata_1.fastq" ) ) in2 = iter( HTSeq.FastqReader( "mydata_2.fastq" ) ) out1 = open( "trimmed_1.fastq", "w" ) out2 = open( "trimmed_2.fastq", "w" ) for read1, read2 in itertools.izip( in1, in2 ): read1.trim_right_end( "ACGGTC" ) read2.trim_left_end( "TTCGAC" ) read1.write_to_fastq_file( out1 ) read2.write_to_fastq_file( out2 ) out1.close() out2.close() I haven't tested the script, and you will have to adjust it to your needs, of course. Note that you can make the trimming tolerant to read errors by either adding a second argument, giving the allowed fraction of mismatches, or using trim_right_end_with_quals, which accepts a maximum sum of mismatch qualities. See here and here for an explanation.

03-28-2012, 09:22 PM	#6
ataheri Junior Member Location: Nashville, TN Join Date: Nov 2011 Posts: 7	Trimmomatic We are using Trimmomatic as it is able to do trimming and adaptor removal for paired-end reads. At first we tried fastx-toolkit but then we switched to trimmomatic as a better choice for paied-end samples.

09-23-2013, 06:21 PM	#7
relipmoc Member Location: Los Angeles, CA Join Date: Jul 2011 Posts: 58	skewer You may also try skewer which is an adapter trimmer dedicated to Illumina paired-end data Last edited by relipmoc; 09-23-2013 at 06:23 PM. Reason: title

06-20-2011, 01:47 AM	#3
fabrice Member Location: paris Join Date: Oct 2009 Posts: 86	Simon Anders, Thanks for your reply. Could you please also give an example how to trim 3' reads which have low quality? For example, set a cutoff value. Bellow this cutoff, the nucleotide will be trimmed. Thanks.

07-20-2011, 06:31 PM	#4
saharnm Junior Member Location: United States Join Date: Jul 2011 Posts: 1	Take a look at "http://solexaqa.sourceforge.net/". It works on both single and paired-end reads.

11-17-2011, 06:40 AM	#5
vbiaudet Member Location: Paris Join Date: Apr 2011 Posts: 13	We have the same problem for paired-end of 100bp and sizing of 200 (fragment size) from Illumina GAIIx RNAseq. In many case we think we got an adapter in 3' that have been sequenced... we think that because the mapping is very bad in 3'. BUT we know that the Quality score is very good (Q> 30). The problem is the 3'end can contains just some bases from the adapter and not the complete sequence... when the list of adapter is known did you find a tool to solve this problem ? Thanks, vb

11-18-2013, 12:54 AM	#8
saran.ela123 Junior Member Location: Singapore Join Date: Aug 2013 Posts: 5	May i know from where to download skewer ?

01-05-2015, 08:48 AM	#9
FlorDV Junior Member Location: Uruguay Join Date: Jan 2015 Posts: 1	Hi saran.ela123 you can download it from here: http://sourceforge.net/projects/skewer/files/