NaOH denaturation with large volumen of low conc library

[madras]

I am doing targeted sequencing using a <15000 kb panel, and because I consequently get rather low library conc (and try to limit the no of PCR cycles) I often use a somewhat large volumen of my library pools for the NAOH denaturation step prior flow cell loading (Nextseq 500 mid).
For rutine library pools the protocol says:
1. Dilute library pool to 2 nM in a 0.1% Tween buffer
2. Denature 5 uL hereof with 5 uL 0.2 M NaOH, neutralize with 5 uL 0.2 M Tris-HCl
3. Do this for all pools to be loaded.
4. Add to each pool HT1 loading buffer to a total volumen (for all pools) of 1 mL.
5. Load onto flowcell.

For the 15k panel:
1. Skip step 1 since the conc is below 2 nM
2. Denature 10 - 50 uL (depending on the conc of the library pool) in 0.2 M NaOH, neutralize with same amount of 0,2 M NaCl.
3-5. Same.

The thing is that it seems as if the cluster generation some times is very inefficient (measured as cluster density per library copies). I suspect that it could have something to do with the larger ratio of (Library+NaOH+Hcl) / HT1 compared to the rutine protocol. In a rutine setup for 4 library pools it would be 4 pools * 5 uL* 3 = 60 uL Library+NaOH+Hcl and 940 uL HT1. The same numbers for 4 40 uL pools in the alternative protocol is 4*40*3 = 360 uL and 640 uL HT1 buffer.

Some concerns could be :
1. A higher amount of Na Cl .
2. Lower molarity of the HT1 constituents
3. Insuffucient buffer capacity with less HT1 (pH is not optimal).

Does anyone have experience with low conc / high vol setup ?

Thanks.
Mads

[gabrieltw]

Try to search older threads. There are a few talking about using low concentration library.

[madras]

Yeah thanks.
I did look through these, e.g. this one. They seem to be somewhat old. My main concern I think is whether HT1 buffer need to be almost "undiluted" - i.e. constitute the majority of the volumen loaded.

[Kodavali]

Hi,

I recently ran Nextseq, I have observed that the cluster density and the total number of reads was very low when compare to the other experiments. Qscore was 97%, Cluster passing filter was above 95% but cluster density K/mm2 was around 94 and the total number of reads for 4 lanes was around 200 million reads. For my previous experiments usually I will get above 180 K/mm2 and around 400 to 500 million reads. I have just followed the same as previous experiments, but not sure where went wrong to get low cluster density and reads.

I have denatured the sample with 0.2 N NaOH, is it the sample that is behaving or something wrong with the preparation?

I used 4nM sample conc with 1.8 loading concentration.

Thanks&Regards,
Kodavali

[keithsweatshirt]

https://bmcgenomics.biomedcentral.com/articles/10.1186/s12864-018-4677-y

I found this paper a couple weeks ago and have been playing around with lower inputs. I've tested with libraries as low as 100 pM on the MiSeq and have gotten good results.

The caveat is that the lower the concentration, the higher the volume of library needed.

These were the steps I did.
1 - 120 uL of 100 pM library
2 - 120 uL of 0.1 M NaOH
3 - Mix and denature for 5 minutes.
4 - Add 240 uL of 0.2 M Tris-HCl pH 7.0
5 - Add 120 uL of HT1 buffer.
6 - Add 20 uL of 20 pM PhiX and then take 600 uL and add to the cartridge.

This ends up being a 20 pM final library and so far we've gotten 900 to 1100 cluster density on the miseq with this. The sequencing quality has looked better too, overall.