Increasing MiSeq v3 cycle number for >600bp reads

JasperGeh · 04-01-2019, 02:07 AM

Hello people!

We have a sequencing application where we could really need, say, 50–100 more cycles from a 600-cycle MiSeq v3 kit.

Illumina states that the cartridge contains reagents for 625 cycles [1] (to account for dual indices we don't need, so 10–15 more cycles there already).
Then again, ECO got about 370 cycles from a 300 cycle kit [2] which is a fair bit more than the 325 stated by Illumina. Other people also mentioned that they pushed above the 25 reserve index cycles [3].

Lastly, this post explains how MadsAlbertsen got 600 cycles from a 400 cycle kit by filling up some wells with reagents from leftover cartridges.

So, my question: Does anybody has experience with increasing the cycle number on the 600-cycle v3 kits and can tell me what safely works? Either by experimenting how many possible cycles are in one cartrige or by filling up from an old cartridge.

Bukowski · 04-01-2019, 03:28 AM

People have been doing similar things on the HiSeq too - but the issue here isn't whether it's technically feasible to hack the kits, it's whether your application will be able to cope with the awful sequencing quality of anything over 300bp in length. If you routinely use the 600 cycle kits, you'll know how the R2 drop off is - so imagine what this is going to be like after 700 cycles...

But see also: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3673215/

JasperGeh · 04-02-2019, 08:02 AM

Thanks, what they did in the paper looks really cool. And yes, I am aware of the decreasing quality, but a small amount of data (after filtering) helps us more than none, in this case...

GenoMax · 04-02-2019, 08:07 AM

Have you considered using nanopore sequencing if you need long(er) reads? No hacking of sequencer needed.

JasperGeh · 04-03-2019, 08:36 AM

Not for this, as it's an antibody library and we need a single paired read with low error rate to cover one fragment. Hacking it will probably be

nucacidhunter · 04-04-2019, 03:12 AM

If you do not need high number of reads, PacBio CCS will be a good option as it will cover the whole fragment length with very high accuracy.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Number of Reads Increasing	hamcan	Bioinformatics	7	01-19-2017 11:27 AM
Increase the number of cycle with a 50 cycles MiSeq Reagent Kit v2	jsvinco	Illumina/Solexa	2	10-21-2015 10:42 PM
different PCR cycle number for different microbiomes?	petralutra	454 Pyrosequencing	1	08-01-2014 06:58 AM
MiSeq 600bp sequencing problem	LizBent	Illumina/Solexa	9	01-30-2014 05:53 AM
Maximum number of cycles from 300-cycle MiSeq kit...	ECO	Illumina/Solexa	13	11-09-2012 04:51 AM

04-01-2019, 02:07 AM	#1
JasperGeh Member Location: Hanover, Germany Join Date: Sep 2018 Posts: 15	Increasing MiSeq v3 cycle number for >600bp reads Hello people! We have a sequencing application where we could really need, say, 50–100 more cycles from a 600-cycle MiSeq v3 kit. Illumina states that the cartridge contains reagents for 625 cycles [1] (to account for dual indices we don't need, so 10–15 more cycles there already). Then again, ECO got about 370 cycles from a 300 cycle kit [2] which is a fair bit more than the 325 stated by Illumina. Other people also mentioned that they pushed above the 25 reserve index cycles [3]. Lastly, this post explains how MadsAlbertsen got 600 cycles from a 400 cycle kit by filling up some wells with reagents from leftover cartridges. So, my question: Does anybody has experience with increasing the cycle number on the 600-cycle v3 kits and can tell me what safely works? Either by experimenting how many possible cycles are in one cartrige or by filling up from an old cartridge.

04-01-2019, 03:28 AM	#2
Bukowski Senior Member Location: UK Join Date: Jan 2010 Posts: 390	People have been doing similar things on the HiSeq too - but the issue here isn't whether it's technically feasible to hack the kits, it's whether your application will be able to cope with the awful sequencing quality of anything over 300bp in length. If you routinely use the 600 cycle kits, you'll know how the R2 drop off is - so imagine what this is going to be like after 700 cycles... But see also: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3673215/

04-02-2019, 08:02 AM	#3
JasperGeh Member Location: Hanover, Germany Join Date: Sep 2018 Posts: 15	Thanks, what they did in the paper looks really cool. And yes, I am aware of the decreasing quality, but a small amount of data (after filtering) helps us more than none, in this case...

04-02-2019, 08:07 AM	#4
GenoMax Senior Member Location: East Coast USA Join Date: Feb 2008 Posts: 7,143	Have you considered using nanopore sequencing if you need long(er) reads? No hacking of sequencer needed.

04-03-2019, 08:36 AM	#5
JasperGeh Member Location: Hanover, Germany Join Date: Sep 2018 Posts: 15	Not for this, as it's an antibody library and we need a single paired read with low error rate to cover one fragment. Hacking it will probably be

04-04-2019, 03:12 AM	#6
nucacidhunter Jafar Jabbari Location: Melbourne Join Date: Jan 2013 Posts: 1,248	If you do not need high number of reads, PacBio CCS will be a good option as it will cover the whole fragment length with very high accuracy.