Question about library normalization methods

[wimverv]

Hi all,

I am looking for a bead based library normalization method as an alternative to the library normalization in the NexteraXT kit.
On the web, I have found commercial kits like the Mag-Bind® EquiPure Library Normalization Kit (Omega Bio-Tek), ALine DNA Normalizer (96 well) v3 (Aline Biosciences or the MagQuant ™ DNA Kit (MagBioGenomics)
I have also found a sort of homebrew protocol, based on AMPure beads (BeNUS protocol; https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4133082/)
Does anyone have experiences with one or more of the above methods, or has an alternative method?
I would like to avoid QPCR measurements as it is the idea to easily upscale/automate the normalization.
Thank you in advance!
Wim

[Yepler]

My group has used ALine kits with good success; I recommend either their v2 or v3 kits. The only caveat is that if you have very low concentration libraries, then you may have more variability or it may not work well. The beads appear to need over-saturation to work well...I don't have the protocol to hand, but I think they suggest 3x input over output for good results.

BeNUS sounded like a *great* idea but didn't work at all for us (my team eventually would hide if they saw me coming with that paper in hand). Possibly it's application-specific and will work better for you.

Hope that helps.

[wimverv]

thank you for your input! Much apreciated.
Wim

[santorini]

I am also looking at these same three kits, and would be very interested in your experiences and results from them.

So far I've tested the Aline Normalization v3 beads and the Mag-Bind Omega beads.

The Aline beads we were very impressed with and got good consistency, even with pretty low concentration samples. Our normal concentration range after Library Prep and Capture isn't very high, so we definitely need the sensitivity on the low end. My only possible complaint is that the beads didn't stick to the magnet very well and I had to be extremely careful I didn't lose beads.

The Mag-Bind beads didn't work well for us because our concentrations were too low. Even after reducing the amount of beads, we never did reach the saturation point of the beads so our end results were all over the place. I think these beads are just meant for very high input amounts (they recommend 2ug). The beads stuck to the magnet very well, but I also changed my magnet for this test. I still have to try the Aline beads with my new magnet to see if that improves the process.

The MagQuant beads I'm still waiting on a sample (out of stock), so I'd be interested to see if you tried them yet.

I've heard from others that highly recommended the Aline beads, so based on their feedback and our promising results I think we'll continue optimizing them.

[Yepler]

MagQuant/Omega beads worked well in my hands if there was plenty of PCR product going into the cleanup.

MagBio beads I never even tested; a chat with a customer service rep there helped me determine that they also required a lot more input than I generally would have. But (this was last spring) the rep did say they were working on a lower-input product.

Hope that helps!

[kgoglin]

I have used Aline beads for normalization in the past with NEBNext libraries. The beads worked well except the final normalized concentration was variable when using different lots. I think Aline has fixed that problem with better QC standards. I'm currently testing them for use with NExteraXT libraries. They seem to require high concentration of libraries (~80nM) in order to get good normalization. Resulting normalized libraries were around 5nM. So we lost a lot of library. Turns out, based on sequencing insert size, the beads were only capturing the smaller fragments. The beads captured about 36% of the total library. We are thinking about increasing the amount of beads to capture more of the library but don't know how that will affect normalization. I believe the Aline Normalization kit is a good option for the NexteraXT libraries, especially because the final normalized library is still double-stranded and can be quantified (not like the normalization beads in the NexteraXT kit which leaves libraries single stranded).

[KimUTA]

Any updated recommendations about the bead based library pooling kits? And kits vs home brews? BTW, anyone else notice the BeNUS paper that the description of index yield in the figure (2) and the text don't match?