Hi Stuart,
Here's what we heard from Illumina about this issue:
1. Expected rate of mis-assigned indexes on a non-patterned flowcell: 0.02%
2. Expected rate of mis-assigned indexes on a patterned flowcell for a
PCR-amplified library stored at -20: <0.5%
3. Expected rate of mis-assigned indexes on a patterned flowcell for a
PCR-amplified library stored at 4 degrees: <0.8%
4. Expected rate of mis-assigned indexes on a patterned flowcell for a
PCR-amplified library stored at RT: <8%
5. Expected rate of mis-assigned indexes on a patterned flowcell for a PCR-free
library: <2%
Bottom-line is we use UDIs for our Kapa libraries but still feel comfortable
running PCR-amplified libraries without UDIs as long as they were properly
stored.
I hope this helps, happy to chat more off-line about our experience so far.
Claire
Rest of post
-----Original Message-----
From: ABRF Discussion Forum <email obscured>> On Behalf Of slevine
Sent: Thursday, May 23, 2019 10:00 AM
To: <email obscured>
Subject: [ABRF Discussion Forum] NovaSeq index hopping
Hello ABRFers,
We've started sending a lot of samples for NovaSeq and the topic of unique dual
indexes (UDIs) come up. Has anyone done any nice studies looking at how much of
an issue this really is? The mantra since
https://www.biorxiv.org/content/10.1101/125724v1 has been to use UDIs on all
ExAmp instruments but Illumina has certainly been working to lower the rate
also. In talking with folks from SeqWell yesterday, in their literature they
say their internal rate is 1:10,000 with their product that does not have UDIs
based on an iSEQ. They compare that to 1:250 for a 'conventional' prep. 1:250
is high, but not awful for most applications certainly, and makes me wonder if
UDIs are strictly necessary for most research applications (obviously going in
informed). This leads to these questions:
1. How well does iSEQ mimic NovaSeq for index hopping?
2. If you do a good job cleaning out free oligo, how much of an issue is index
hopping on NovaSeq with routine applications (NextFlex, Kapa, SMARTseq, etc.)
-Stuart
――
View topic http://list.abrf.org/r/topic/6p7C8hMsdbCKuKhhhDINHK
Leave group <email obscured>?Subject=Unsubscribe