In Section 1.5. Lysis and Sublibrary Generation of the barcoding protocol, the pool of cells/nuclei is counted and split into what we call sublibraries.
At this point, each cell/nucleus has received the first three barcodes. You can choose the number of cells/nuclei added per sublibrary. The recommended maximum number of cells/nuclei loaded per sublibrary will depend on the kit size:
| Kit size | Max output | # sublibraries | # cells/nuclei per sublibrary |
| Mini | 10,000 | 2 | 5,000 |
| WT | 100,000 | 8 | 12,500 |
| Mega | 1,000,000 | 16 | 62,500 |
| Penta | 5,000,000 | 32 | 156,250 |
After splitting the cells/nuclei into each sublibrary, the cells/nuclei are lysed, and a unique 4th barcode, i5/i7 unique dual index (UDI), is added in Section 3.5. Barcoding Round 4. After sequencing, the Parse Biosciences Analysis Pipeline assigns reads that share the same 4 barcode combination to a single cell/nucleus.
Sublibraries can be sequenced together or separately. One strategy to QC sample and library quality is to sequence one sublibrary alone at a sequencing depth above the minimum of 20,000 reads/cell. These results could also help decide the sequencing number of reads required for the remaining sublibraries to achieve the desired number of transcripts and genes per cell.