Skip to content

QuantumScale Demultiplexing

The Core Concepts:

1. Library level, NOT sample level FASTQs

The FASTQ files generated at this stage represent library-level demultiplexing based on sub-libraries (PCR primer pools (i5) ) and bead barcode groups (i7), not biological samples. Each FASTQ file contains reads from multiple different biological samples that share the same library-level barcodes. The actual sample-level separation occurs later in the ScaleRNA pipeline when the workflow processes the cell barcodes and RT barcodes to assign reads to specific biological samples according to the sample barcode table configuration.

2. Partial Bead Barcode Demultiplexing

QuantumScale uses a demultiplexing strategy that manages file organization and processing efficiency. FASTQ generation uses only the first 8bp of the 32bp bead barcode, creating 96 FASTQ file sets per PCR pool, while preserving the complete 32bp sequence for downstream analysis. This approach reduces the total number of FASTQs from over 1 million per PCR pool to 96 FASTQ file sets (file set means R1, R2, I1, I2; 4 total files) per PCR pool.

Key Strategy Components:

  • Demultiplexing: First 8bp of 32bp bead barcode used for FASTQ file organization
  • Preservation: Complete 32bp sequence maintained for downstream analysis
  • Efficiency: 96 FASTQ file sets per PCR pool instead of millions of individual files

Why This Approach?

Problem: QuantumScale bead barcodes contain over 1 million unique index combinations per PCR pool. Creating separate FASTQ files for each would be computationally impractical.

Solution: Use the first 8bp of the 32bp bead barcode for demultiplexing, while preserving the complete 32bp sequence for downstream analysis.

Key Components Explained

1. Bead Barcode (i7) - The 32bp Strategy with 8bp Demux

Full 32bp Bead Barcode: GTATCCTTCTGTCAGCTAACGGTTGCATGTTA
                        ├──8bp─┤├─────────24bp─────────┤
                        Demux   Additional
                        Portion  Sequence

What happens:

  • First 8bp: Used for FASTQ demultiplexing (creates 96 unique fastqs per QSR index)
  • Full 32bp: Preserved for complete downstream analysis by Scale Bio Seq Suite

2. Sample_ID Strategy - Merging

Instead of creating separate files for every possible bead barcode combination, the strategy creates 96 file sets per PCR pool by using identical Sample_IDs for different bead barcode combinations.

Example:

Sample_ID: QSR-1_001
  • Contains reads from multiple different bead barcodes
  • All sharing the same first 8bp sequence
  • Merged into single FASTQ file for parallel processing

FASTQ Read Structure Examples

Read 1 (RNA Sequence)

@INSTRUMENT:LANE:FLOWCELL:1:TILE:X:Y:CTGTCAGCTAACGGTTGCATGTTA 1:N:0:GTATCCTT+TCGATGAT
GCACGAGCTGCACCGGGGCCCGCGCAGCAGCAGGGCCCTGCGGCCTGCCAGCATGGATCTCCTGCGCCCACACTGGCTGGAG
+
IIIIIIIIIIIIIIIIIII9IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII9IIIIIIIIIIIIIIIIIIIIIIIIIIIIII

Read 2 (RT Barcode + Molecular Barcode)

@INSTRUMENT:LANE:FLOWCELL:1:TILE:X:Y:CTGTCAGCTAACGGTTGCATGTTA 2:N:0:GTATCCTT+TCGATGAT
GTCGTGTCGGAGATTG
+
I9IIIIIIIIII-III

Index 1 (i7 - Bead Barcode - Full 32bp Preserved)

@INSTRUMENT:LANE:FLOWCELL:1:TILE:X:Y:CTGTCAGCTAACGGTTGCATGTTA 1:N:0:GTATCCTT+TCGATGAT
GTATCCTTCTGTCAGCTAACGGTTGCATGTTA
+
IIII99IIIIIII9I9III9IIIIIIIIIIII

Note: GTATCCTT (first 8bp) is used for demultiplexing, but the complete 32bp sequence is preserved for analysis

Index 2 (i5 - PCR Primer Pool)

@INSTRUMENT:LANE:FLOWCELL:1:TILE:X:Y:CTGTCAGCTAACGGTTGCATGTTA 2:N:0:GTATCCTT+TCGATGAT
TCGATGAT
+
I9IIIIII

Common Questions Answered

Why are there 1-96 unique Sample_ID names for each QSR PCR index?

Answer: The 96 Sample_IDs reduces the total number of FASTQ files 4-fold because the read data will be exported together for each unique Sample_ID. The full 32bp bead barcode information is still preserved in each FASTQ file for complete downstream analysis.

Why are there 4 unique i5 indexes for each unique i7?

Answer: Each QSR PCR index pool contains 4 different primer sequences to enable sub-library identification and parallel processing. The i5 (Index 2) read contains PCR primer pool information that distinguishes between different sub-libraries within the same experiment. The 4 primers per pool provide required diversity for sub-library identification. For example:

  • Large/XL kits: 8 PCR pools × 4 primers = 32 total i5 sequences
  • Modular kits: 12 PCR pools × 4 primers = 48 total i5 sequences

Why are some Sample_ID's the same?

Answer: Instead of creating separate files for every possible bead barcode combination (which would generate millions of files), the strategy creates exactly 96 FASTQ file sets per PCR pool by using identical Sample_IDs for different bead barcode combinations that share the same first 8bp sequence.

Important

All ScaleBio libraries require these 4 fastq files:

  • Read 1 (R1)
  • Read 2 (R2)
  • Index 1 (I1, i7)
  • Index 2 (I2, i5)

QuantumScale Samplesheets

All QuantumScale samplesheets are available in the QuantumScale samplesheets.

Need Help?

For more information, please contact support@scale.bio or visit our support website.