How to pool NGS libraries at equal molarity
When several sequencing libraries share a run, each one should get a fair share of the reads. Sequencers count molecules, not nanograms, so libraries are pooled by molarity. This guide walks through the calculation with a real example.
Why pool by molarity, not mass
A nanogram of short fragments contains more molecules than a nanogram of long fragments. If you pool equal masses of a 300 bp library and a 600 bp library, the shorter one contributes about twice as many molecules and takes about twice as many reads. Converting everything to nM first removes that bias.
Step 1: Measure concentration
Use a fluorometric dsDNA assay such as Qubit. Absorbance (NanoDrop) also counts free nucleotides, primers and single-stranded DNA, so it usually overestimates library concentration.
Step 2: Measure the average fragment size
Run each library on a Bioanalyzer or TapeStation and read the average size of the library peak, including adapters. Exclude primer-dimer or adapter-dimer peaks from the region.
Step 3: Convert ng/µL to nM
Use an average mass of 660 g/mol per base pair:
nM = ng/µL × 10⁶ ÷ (660 × average size in bp)
Example: a library at 12.4 ng/µL with an average size of 420 bp is 12.4 × 10⁶ ÷ (660 × 420) = 44.7 nM.
Step 4: Work out each library's volume
Decide the final pool concentration and volume. Each of n libraries contributes an equal share of the pool's molarity, so each library's volume is:
V(library) = (pool nM ÷ n) × pool volume ÷ library nM
For a 4 nM, 50 µL pool of four libraries, each library should end up at 1 nM:
| Library | ng/µL | Size (bp) | nM | Volume (µL) |
|---|---|---|---|---|
| Lib_01 | 12.4 | 420 | 44.7 | 1.12 |
| Lib_02 | 8.1 | 390 | 31.5 | 1.59 |
| Lib_03 | 21.7 | 450 | 73.1 | 0.68 |
| Lib_04 | 5.3 | 410 | 19.6 | 2.55 |
| Water or EB | 44.06 |
Practical tips
- Avoid sub-microlitre volumes. Pipetting under about 0.5 µL is inaccurate. Pre-dilute concentrated libraries (for example 1:10) or make a larger pool.
- Want unequal read depth? Give a library a larger share of the pool molarity. A sample that needs twice the reads gets twice the share.
- Confirm the pool by qPCR. Only molecules with both adapters form clusters. A library quantification qPCR on the final pool gives the most accurate loading concentration.
- Then follow your instrument's guide for denaturation and final loading concentration, which differ between platforms and kits.
Frequently asked questions
Why use 660 g/mol per base pair?
It is the commonly used average molar mass of one double-stranded DNA base pair as a sodium salt. Different sources use values from about 617 to 660; the choice shifts all libraries equally, so equimolar pooling is unaffected.
Should I use Qubit or NanoDrop for library concentration?
Qubit or another fluorometric dsDNA assay. Absorbance-based readings include free nucleotides, primers and single-stranded DNA and tend to overestimate library concentration.
Can I pool libraries of different sizes?
Yes, as long as you pool by molarity rather than mass. Very different sizes can still cluster with slightly different efficiency, so check read distribution after the first run and adjust.