Explanation on Paired end single index for RNA seq

Question

[hothriananya]

Hi,

I have been doing RNA seq analysis recently, my reads are paired end but when I am doing bowtie the results are coming up as 100% unpaired. I have tried different ways for aligning and I am getting the same answer. My question is does the bowtie2 give results as unpaired if the run is paired end single index and paired if the run is paired end mutli index?

Thanks in advance

[Devon Ryan]

Show the exact command that you're using. You're probably just using the wrong options. Also, are you sure you want to use bowtie2 for RNAseq alignment? That's only appropriate if you're aligning to the transcriptome or if your organism lacks splicing.

[hothriananya]

As I have so many reads per sample

bowtie2 -q mm9 \
  -U Sample1.fastq, Sample2.fastq \
  -U Sample3.fastq,Sample4.fastq \
  -U Sample5.fastq,Sample6.fastq \
  -U Sample7.fastq,Sample8.fastq \
  -S Sample.sam

Also I have combined the sample and tried the following command too, but I got the same answer,

bowtie2 --very-sensitive mm9 gfp.fastq -S gfp.sam

[Devon Ryan]

You're explicitly telling bowtie2 that you have single-end reads. Also, you can't give multiple -U options and have any hope that bowtie2 (or any other command) will be able to discern what you want it to do. Are Sample1, Sample2, etc. actually different samples that you want in separate files or are they simply technical replicates of a single sample?

[hothriananya]

They are replicates of single sample. I am a beginner to RNA seq and trying to learn. Could you help me with the syntax please.

[Devon Ryan]

Then you need to seperate all of them by commas:

bowtie2 mm9 -U Sample1.fastq,Sample2.fastq,Sample3.fastq,Sample4.fastq

You're still telling bowtie2 that you have single end reads, of course. If you actually do have paired end reads, then you should have 2 files for each replicate. You then use the -1 and -2 options rather than -U.

[hothriananya]

I even used the above option u mentioned and it returned with a error,

Error: Encountered internal Bowtie 2 exception (#1)
Command: /usr/bin/bowtie2-align --wrapper basic-0 -q -3 S5L5_S1_L003_R1_001.fastq,S5L5_S1_L003_R2_001.fastq mm9 -4 S5L5_S1_L004_R1_001.fastq,S5L5_S1_L004_R2_001.fastq -S Wild1.sam -1 S5L5_S1_L001_R1_001.fastq,S5L5_S1_L001_R2_001.fastq -2 S5L5_S1_L002_R1_001.fastq,S5L5_S1_L002_R2_001.fastq
bowtie2-align exited with value 1

This is the command I have used

bowtie2 -q mm9 \
  -1 S5L5_S1_L001_R1_001.fastq,S5L5_S1_L001_R2_002.fastq \
  -2 S5L5_S1_L002_R1_001.fastq,S5L5_S1_L002_R2_001.fastq \
  -3 S5L5_S1_L003_R1_001.fastq,S5L5_S1_L003_R2_001.fastq \
  -4 S5L5_S1_L004_R1_001.fastq,S5L5_S1_L004_R2_001.fastq \
  Wild1.sam

[Devon Ryan]

There are no such options as -3 or -4.

[Ashutosh Pandey]

+1 for Devon's patience :-)

[hothriananya]

Definitely a +1 to Devon's patience, at least he is humble and helpful rather passing a sarcastic comments.

[hothriananya]

I have even tried doing like this

bowtie2 -q mm9 \
  -1 S5L5_S1_L001_R1_001.fastq,S5L5_S1_L001_R2_001.fastq,S5L5_S1_L002_R1_001.fastq,S5L5_S1_L002_R2_001.fastq \
  -2 S5L5_S1_L003_R1_001.fastq,S5L5_S1_L003_R2_001.fastq,S5L5_S1_L004_R1_001.fastq,S5L5_S1_L004_R2_001.fastq \
  -S WT_1.sam

and it is giving me error as below:

Error, fewer reads in file specified with `-1` than in file specified with `-2`

terminate called after throwing an instance of 'int'
bowtie2-align died with signal 6 (ABRT) (core dumped)

[Devon Ryan]

Ah, a real example! That's helpful. Here is what you meant to type:

bowtie2 mm9 -1 \
S5L5_S1_L001_R1_001.fastq,S5L5_S1_L002_R1_001.fastq,S5L5_S1_L003_R1_001.fastq,S5L5_S1_L004_R1_001.fastq \
-2 S5L5_S1_L001_R2_001.fastq,S5L5_S1_L002_R2_001.fastq,S5L5_S1_L003_R2_001.fastq,S5L5_S1_L004_R2_001.fastq \
-S WT_1.sam

Note that all of the R1 samples are after -1 and all of the R2 samples are after -2. Notice also that the orders of the pairs between -1 and -2 are identical. Doing this in any other way would lead to useless results.

Note, I've made the command multiple lines with "\" for legibility.

[hothriananya]

Thank you very much, It worked :)

[gopal_joshii]

Does overlaping paired-end reads help in reference based assembly?

Hi All, I want to resequence a bacterial genome and do reference based assembly. As per the available literature, I came to know that Illumina NextTerra protocal uses tagmentation for DNA fragmentation and generate random sequences of ~300 nt length. In this case, what should be my sequencing layout 2x250 paired end or 2x150 paried-end sequencing. As I have to generate reference based assembly does this overlapping 2x250 will provide extra edge or 2x150 will give be enough as it will cover 300 nt insert? Please provide your valuable suggestions so that I can plan my experiment accordingly. Thanking you.

[lakhujanivijay]

Dear gopal_joshii

This is an unrelated post in context to the original question and should a treated and posted as a new question. This will help others and you as well.

At biostars, we recommend re-posting this as a new question after making sure that similar question has not been asked previously.

Thanks
Vijay