Assess the assembly quality using unpaired reads

[wes]

Dear All

After trimming RNAseq dataset using Trimmomatic, there are paired-end reads and unpaired reads (consist of both forward and reverse unpaired reads). 

I only used the paired end reads to proceed with de novo assembly. To check the output of the assembly quality, bowtie2 was run by using paired-end reads. Overall the alignment rate is more than 90%. 

However, I would like to assess the quality of the assembly by using unpaired reads. After running the bowtie2 using the unpaired reads, I found out that the overall alignment rate is high for the forward unpaired reads (more than 90%) but significantly low for the reverse unpaired read ( less than 15%). May I know why?

[tiag...@me.com]

Can you post the bowtie command you used to map the unpaired reads? My guess is that you used the wrong strandness.

 

I don’t quite see why you would use the unpaired data; it will not tell you anything the pair data hasn’t told you.

 

If you want more metrics, use ExN50, BUSCO and transrate (or DETONATE).

 

T.

--
You received this message because you are subscribed to the Google Groups "trinityrnaseq-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/trinityrnaseq-users/9b84684e-fd77-4712-a927-c1b0d4b1028bn%40googlegroups.com.

[wes]

This is the command 

bowtie2-build Trinity.fasta Trinity.fasta  

bowtie2 -p 10 -q --no-unal -k 20 -x Trinity.fasta -U SRR5904767_1.unpaired.fastq.gz 2>align_stats_UP_TA1_1.txt| samtools view -@10 -Sb -o bowtie2_UP_TA1_1.bam --threads 32

bowtie2 -p 10 -q --no-unal -k 20 -x Trinity.fasta -U SRR5904767_2.unpaired.fastq.gz 2>align_stats_UP_TA1_2.txt| samtools view -@10 -Sb -o bowtie2_UP_TA1_2.bam --threads 32

bowtie2 -p 10 -q --no-unal -k 20 -x Trinity.fasta -U SRR5904768_1.unpaired.fastq.gz 2>align_stats_UP_TA2_1.txt| samtools view -@10 -Sb -o bowtie2_UP_TA2_1.bam --threads 32

bowtie2 -p 10 -q --no-unal -k 20 -x Trinity.fasta -U SRR5904768_2.unpaired.fastq.gz 2>align_stats_UP_TA2_2.txt| samtools view -@10 -Sb -o bowtie2_UP_TA2_2.bam --threads 32

The reason of using unpaired read is just to assess the quality of the assembly on top of normal checking parameter ExN50, BUSCO and transrate

[tiag...@me.com]

I can’t particularly see from this why the reverse reads are not mapping, if the pairs did. I also still don’t understand why you think you will getting anything more from unpaired reads. The paired mapping is a much better metric than unpaired, as it takes concordancy into account, and potentially strandness. The unpaired mapping is literally a worse metric than the paired mapping. So much so that before bowtie2 we used a perl script to analyze unpaired mappings as if they were paired.

To view this discussion on the web visit https://groups.google.com/d/msgid/trinityrnaseq-users/87be970d-a447-4f1c-9ba8-741b8155fc08n%40googlegroups.com.