Reply
Highlighted

Results from different HiSeqs

Posted By: graham, on Sep 29, 2016 at 8:08 AM

Hi,

I was looking at the Sample Requirements & Library Considerations  page and noticed that when using 150bp reads from the HiSeq 4000, assemblies produced contigs half as long as those from HiSeq X.

Does anyone know why this is?

Also, does anyone have experience of running 10x libraries on the HiSeq 2500 in high-output mode? This is listed as not tested on the link above.

Any comments are most gratefully received.

Cheers,

Graham

Dr. Graham Etherington
The Earlham Institute, Norwich, UK
Twitter: @bioinformatiks

3 Replies

Re: Results from different HiSeqs

Posted By: jafar, on Sep 30, 2016 at 6:33 PM

Sequence quality on HS 2500  HO 2x125 was very good and comparable to shotgun libraries. With HS 2500 shorter read less proportion of data are used for assembly or phasing in comparison to HS X 2x150 since 16 bp of each read pair is the barcode and 6-8 seems to be randomers which are not included in the assembly. Trimming NA12787 HS X data to 125 bp reduced contig and phased blocked length slightly.

 

Re: Results from different HiSeqs

Posted By: graham, on Oct 3, 2016 at 6:06 AM

Thanks for that. Any idea why HS 4000 performs so badly when using 150bp reads?

Dr. Graham Etherington
The Earlham Institute, Norwich, UK
Twitter: @bioinformatiks

Re: Results from different HiSeqs

Posted By: james, on Nov 22, 2016 at 1:28 AM

Hi Graham,

Is it just the contig length that's affected? Or are other aspects of the sequence reads bad too?

We've only run a few libraries and all on HiSeq 2500, and performance was OK. There does not seem to be a good place to report run metrics for 10X so users can share experiences (perhaps you could knock something up). The different clustering chemistry on 2500/4000 would have an impact. But I am not sure why the X would perform so much better than the 4000. Was this exactly the same library?

James.