paired end result for preprocessing

Question

[samh]

Hi,

I am doing preprocessing analysis for RNA-seq. After I removed the adapter, the total sample reads is still the same as total high quality reads. The worse problem is, when I run the select paired end read, the result obtained was zero bytes. So how can I overcome this problems?

Thank you.

[Devon Ryan]

"when I run the select paired end read"

Can you clarify what you mean by that? I understood that trimming didn't remove any reads (that's good, it just means you had minimal adapter contamination), but I haven't a clue what you mean by what I quoted above.

[samh]

What I mean by that quoted is, I want to obtain paired end reads after remove adapter sequence. But in my case, the paired end reads that I obtained give me no result. So, as you know, if we can't identify the paired end reads, we can't proceed to get shuffled paired end sequence.

[Devon Ryan]

How are you trimming them? They should be trimmed as pairs, or are you doing that and everything is being removed?

[samh]

I had trimmed them separately based on base quality (Qphread=20) and sequence length (min 50 bp). I had done these step in previous analysis with different data and it works. However, I am having problem with current analysis. For your additional information, the size of my reads is up to 5gb each while my previous size of reads is only up to 1gb. It is the size of the reads will influence the outcome for paired end reads?

[h.mon]

Could you post which tool and the commands you are using?

[samh]

I only use perl script command line for my preprocessing analysis.

This is example how I trimmed the reads based on base quality (Qphread = 20)

DynamicTrim.pl -h 20 reads_R1.fastq
DynamicTrim.pl -h 20 reads_R2.fastq

where reads_R1 & reads_R2 are referring to forward and reverse file.

[samh]

I have one additional question, now I'm trying to use trimmomatic tool to trim the reads. However, based on the manual that was provided, I need to insert trim log file to run the command. But, I don't know how to create trim log file.

This is the command that I used:

java -jar <path to trimmomatic.jar> PE [-threads <threads] [-phred33 | -phred64] [-trimlog <logFile>] >] [-basein <inputBase> | <input 1> <input 2>] [-baseout <outputBase> | <unpaired output 1> <paired output 2> <unpaired output 2> <step 1>

[Devon Ryan]

You don't create the log, the tool does and puts it into a file whose name you provide. Note that the log is optional, so you don't need to provide it.

[samh]

Thank you Devon Ryan for the info. I will try again.

Answer 1

[h.mon]

By quality-trimming them independently, you have broken read pairing. You should use any program which keeps pairing for trimming (BBDuk, cutadapt, SeqyClean, many others).

You may also use Repair (from BBTools) to fix the pairing from the reads trimmed by DynamicTrim.

[5heikki]

I've been using this script for read pair resynchronization.

[samh]

thank you for sharing this script :)

[samh]

I'm quiet confused about why we should trimmed read as pairs. I already check the quality for both reads via fastqc tools and it shows my reads have high quality scores. My problem now is when I want to select the paired end read that remained after I had removed all the contaminant such as adapter, the result shown zero byte means there is nothing in the file.

[Devon Ryan]

We should trim reads as pairs to avoid the annoyance of resyncing them again.

[h.mon]

Paired reads add information, they should follow an expected insert size distribution and have proper orientation. This information is used by assemblers and mappers to increase accuracy, and is helpful troubleshooting problems.