Optimal Cluster Density
The density of clusters on a flow cell significantly impacts data quality and yield from a run, and is a critical metric for measuring sequencing performance. It influences run quality, reads passing filter, Q30 scores, and total data output.
Performing a run at optimal cluster density involves finding a balance between underclustering and overclustering. The goal is to sequence at a high enough density to maximize total data output, while maintaining a low enough density to avoid overclustering.
Effects of Underclustering and Overclustering
Overclustering increases signal brightness, which makes finding the focal plane difficult and causes poor template generation, poor cluster registration, and other image analysis issues. These issues negatively affect sequencing data in the following ways:
| • | Lower Q30 scores—Overloaded signal intensities decrease the ratio of base intensity to background, creating ambiguity during base calling and decreasing data quality. |
| • | Lower clusters passing filter (lower data output)—Overclustered flow cells typically have more overlapping clusters, which cause poor template generation and a decrease in percent of clusters passing filter (%PF). The %PF metric indicates signal purity from each cluster. Lower %PF reduces yield (the number of bases in gigabases [Gb]) called for a run. |
| • | Inaccurate demultiplexing—Index reads typically have lower diversity, which can cause poor base calling. Overclustering exacerbates the potential for poor base calling, leading to demultiplexing failure. |
| • | Run failure—When overclustering is extreme, image focusing can fail and terminate the run at any cycle. |
Underclustering maintains high data quality, but lowers data output. In general, underclustering is preferable to overclustering because the effects are less severe.
Recommended Cluster Densities
When targeting optimal cluster density for nonpatterned flow cells, use the raw cluster density range for your system and reagent kit as a guideline.
|
System |
Reagent Kit |
Raw Cluster Density (K/mm²) |
|---|---|---|
|
HiSeq 2500 (High Output) |
HiSeq v4 |
950–1050 |
|
TruSeq v3 |
750–850 |
|
|
HiSeq 2500 (Rapid Run) |
HiSeq v2, TruSeq (v1), and Rapid Duo |
850–1000 |
|
MiniSeq |
MiniSeq High Output and Mid Output |
170–220 |
|
MiSeq |
MiSeq v3 |
1200–1400 |
|
MiSeq v2 |
1000–1200 |
|
|
NextSeq |
High Output and Mid Output (v2.5 and v2) |
170–220 |
Density is measured as 1000 (K) clusters per square millimeter (mm²). Raw cluster density indicates how many clusters are on the flow cell, regardless of whether they passed filter.
Raw cluster density is not a useful metric for pattered flow cells because the ordered arrangement of nanowells ensures uniform cluster density.
