high levels of broken read pairs?

[gevielr]

I've been trying to run de novo transcriptome assemblies in CLC, but I am getting high levels of broken pairs (~65%). I am wondering if this is normal?
My insert sizes were 200-250 bp, and so I set my minimum and max read distance as 150-300. I've manipulated this number, but it only made it worse.

I have ~140M paired reads.

I'm wondering if I did something wrong with processing? I did quality-based and adapters trimming and based on quality using fastq-mcf, which is meant for paired reads (discards reads as pairs), everything looks OK when I view them in CLC (correctly paired etc...)

[asteraceae]

Hi there,

I am running across a similar problem with my analysis in CLC. I am wondering if you resolved your problem?

I have ~70Gb of illumina PE data (101bp with ~insertsize of 250bp)

I ran the assembly with the auto detect paired distances option. I have included my summary file here in case that helps.

Thanks in advance