Nextera XT w/ metagenomes of various sizes

[adam.geber]

I'm planning to make libraries from several metagenomic DNA samples using Nextera XT for the purpose of de novo assembly of microbial genomes. I'm unconcerned with complete reconstruction of individual genomes but as I would like to take a "gene-centric" approach to these metagenomes I want to maximize depth to aid in assembly. My question is: given DNA input of constant mass but from samples of very different fragment size distributions (see attached Tapestation traces) how consistent can I expect my library insert sizes to be? I assume that if Nextera chemistry is used for both amplicon and whole genome sequencing that Tn5 activity must be influenced more by template mass than anything but I wanted to ask the community.

Thank you!

[nucacidhunter]

Generally, library profile down to 3 kb fragment will be similar for similar input quantity. Inputs shorter than 3 kb will result in libraries with shorter average insert size but one can remove smaller library fragments depend on sequencing cycles to minimise R1 and R2 overlap if it is desirable. Your samples in D1, E1 and F1 would give similar profiles but other two will have more small inserts. You may prepare two libraries from those two, remove smaller fragments and combine them for sequencing to maximise diversity.

[adam.geber]

Thanks for your suggestions, nucacidhunter! I went with half-volumes of the Nextera XT system (halving the input mass as well and performing one extra cycle of PCR to compensate i.e. 13 total), which resulted in very consistent insert sizes following a 0.5X AMPure bead cleanup. I've attached a gDNA screentape trace with 6 library pools (each ~1kb on average) and 2 RNA-seq libraries (prepared by the NEBNext Ultra RNA kit) for reference. These are intended for a 2x250 HiSeq Rapid Run -- I'll report back if there's anything worth sharing.