Insert size different from company with BBMap

Question

[Sharon]

Hi All

Insert size in BBMap is different than what the company mentioned? Which is correct? We got ~150, company says its 200 ! How can we have reads of 150 each and inner mate -135 and insert size is 200. Insert size - 2 * read length = inner mate then if inner mate is -135 and insert is 200, then, we have double read = 335 , then read is longer than 150, while we have 15o bases in the fastq for each file.

Thanks

Answer 1

[GenoMax]

BBMap. Since it is based on actual merging of reads and/or alignment data. Not sure which way you did the analysis ( C: Target fragment size versus final insert size ).

[Sharon]

Sorry, I made update before your comment. So we can't have inner mate with -135 and insert = 200 and read length is 150 in fastq file , right? I used this:

bbmerge.sh in1=r1.fq in2=r2.fq ihist=ihist.txt

[GenoMax]

Can you restate your question referring to this image (A: What is the different between Read and Fragment in RNA-seq? )? If the reads overlap then then is no inner distance (ref fig).

[Sharon]

I think if the reads overlap, we will have negative inner mate, right? And if the insert size is shorter than the sum of the 2 reads then the reads will overlap, right? Like if we have 150 bases in each fastq file, and we have insert size as claimed by the company 200, and the inner mate as claimed by the company is -135, then the numbers don't sum.

insert size - 2 *read length = inner mate

200 - ~300 != -135

but if the insert size is 158 as by bbmap, then we can have :

158 - ~300 = ~-142 right?

[GenoMax]

Is the company counting the adapters in the fragment length they are quoting?

If BBMap estimates the fragment size to be 158 then it means that you are sequencing 150 bp from end #1 leaving a 8 bp piece on the right. When R2 comes along, it is going to cover this 8 bp piece first but will leave a corresponding 8 bp fragment on the left side of the fragment.

|-------------------------------->8 bp|    R1 - 150 bp
|-------------------------------------|    Fragment 158 bp
|8 bp<--------------------------------|    R2 - 150 bp

[Sharon]

I will check what the company mean. But does BBmap estimate fragment size or insert size? And can I trust BBMap and argue with company based on BBMap estimation?

[GenoMax]

Insert size. Fragment size is insert size + length of 2 adapters. If you have a reference available then you should be able to see this clearly if you use the mapping method. Map only a million or so reads.

[Sharon]

Thanks genomax so much. I think it is clear now, BBmap estimates makes more sense !