What are P5 and P7 adapter sequences?

P5 (5'-AATGATACGGCGACCACCGAGA-3') and P7 (5'-CAAGCAGAAGACGGCATACGAGAT-3') are universal sequences that bind to complementary oligos grafted on the Illumina flow cell surface. They are essential for cluster generation via bridge amplification. P5 is located at the read 1 end, P7 at the read 2 end. These sequences are identical across all Illumina library preparation kits.

What is the difference between combinatorial and unique dual indexing?

Combinatorial dual indexing reuses individual i5 and i7 index sequences across samples — uniqueness comes from the combination (e.g., i7-A + i5-1 vs i7-A + i5-2). If index hopping occurs, a hopped read could match a valid but wrong sample combination. Unique dual indexing (UDI) assigns a globally unique i5+i7 pair per sample — if hopping occurs, the resulting pair matches no valid sample and is filtered as "undetermined." UDI is strongly recommended for patterned flow cells.

Why does index hopping occur on NovaSeq but not MiSeq?

Index hopping is predominantly associated with Exclusion Amplification (ExAmp) chemistry used on patterned flow cells (NovaSeq 6000, NextSeq 2000, NovaSeq X). In ExAmp, DNA fragments and amplification reagents coexist in solution on the flow cell, allowing free-floating adapter molecules to mis-prime nearby clusters. Older platforms (MiSeq, HiSeq 2500) use bridge amplification on random flow cells, where this mechanism is much less likely. Typical hopping rates: 0.1-2% on ExAmp platforms vs <0.01% on non-patterned platforms.

How do I design custom Illumina-compatible adapters?

Custom adapters must preserve: (1) P5/P7 sequences exactly for flow cell binding, (2) sequencing primer binding sites (SP1 for Read 1, SP2 for Read 2), (3) correct orientation of i5/i7 indexes relative to the sequencing read direction for your platform. Validate Tm of the double-stranded adapter region (target 55-65°C at 50 mM NaCl), check for self-complementarity (ΔG > −6 kcal/mol), and verify no hairpins in single-stranded overhangs (ΔG > −2 kcal/mol). Use our Tm Calculator and Secondary Structure Predictor for validation.

What causes adapter-dimer contamination and how do I prevent it?

Adapter dimers form when two adapter molecules ligate to each other instead of flanking a DNA insert. They are ~120-130 bp (TruSeq) and sequenceable, wasting reads. Prevention: (1) use optimal adapter:insert molar ratio (typically 10:1 to 25:1), (2) use SPRI bead cleanup at 0.8x ratio to size-select away dimers, (3) for low-input libraries (<1 ng), reduce adapter concentration by 10-fold, (4) validate by Bioanalyzer/TapeStation — dimers appear as a sharp peak at ~120-130 bp.

Which indexing strategy should I use for my experiment?

Use UDI (Unique Dual Indexing) whenever possible — it is the Illumina recommendation for all patterned flow cell platforms (NovaSeq, NextSeq 2000, NovaSeq X). Use combinatorial indexing only on non-patterned platforms (MiSeq, iSeq) where index hopping is negligible. For custom panels with 96+ samples, pre-plated UDI kits (e.g., Illumina DNA/RNA UD Indexes Set A-D, 384 unique pairs) are the most practical option. Never mix UDI index sets from different product versions in the same sequencing run.

How to Design Illumina Adapters and Dual Indexes for NGS Library Prep

Q: How do I design custom Illumina-compatible adapters?

Custom adapters must preserve: (1) P5/P7 sequences exactly for flow cell binding, (2) sequencing primer binding sites (SP1 for Read 1, SP2 for Read 2), (3) correct orientation of i5/i7 indexes relative to the sequencing read direction for your platform. Validate Tm of the double-stranded adapter region (target 55-65°C at 50 mM NaCl), check for self-complementarity (ΔG > −6 kcal/mol), and verify no hairpins in single-stranded overhangs (ΔG > −2 kcal/mol). Use our Tm Calculator and Secondary Structure Predictor for validation.

Q: Which indexing strategy should I use for my experiment?

Use UDI (Unique Dual Indexing) whenever possible — it is the Illumina recommendation for all patterned flow cell platforms (NovaSeq, NextSeq 2000, NovaSeq X). Use combinatorial indexing only on non-patterned platforms (MiSeq, iSeq) where index hopping is negligible. For custom panels with 96+ samples, pre-plated UDI kits (e.g., Illumina DNA/RNA UD Indexes Set A-D, 384 unique pairs) are the most practical option. Never mix UDI index sets from different product versions in the same sequencing run.

1. Which Parts of an Illumina Adapter Must Stay Fixed?

A fully adapted Illumina library molecule has the following structure (5' → 3' of the top strand):

// TruSeq / Illumina DNA Prep fully-adapted molecule

5'- [P5] — [SP1] — [i5 index (8 bp)] — [SP2] — [INSERT] — [SP2'] — [i7 index (8 bp)] — [SP1'] — [P7] -3'

// Functional regions:

P5: 5'-AATGATACGGCGACCACCGAGA-3' (22 nt — flow cell binding)

P7: 5'-CAAGCAGAAGACGGCATACGAGAT-3' (24 nt — flow cell binding)

SP1: Sequencing primer 1 binding site (Read 1 initiation)

SP2: Sequencing primer 2 binding site (Read 2 initiation)

i5: Index 2 (8 bp barcode) (demultiplexing)

i7: Index 1 (8 bp barcode) (demultiplexing)

Why must P5 and P7 stay unchanged?

P5 and P7 are universal sequences complementary to oligos covalently attached to the flow cell surface. During cluster generation, denatured library molecules hybridize to these surface oligos, initiating bridge amplification (random flow cells) or exclusion amplification (patterned flow cells). These sequences are invariant across all Illumina library preparation kits — TruSeq, Nextera, Illumina DNA Prep, and custom protocols must all preserve P5/P7 exactly.

Sequence	5' → 3'	Length	GC%	Approx. Tm
P5	AATGATACGGCGACCACCGAGA	22 nt	54.5%	~67°C
P7	CAAGCAGAAGACGGCATACGAGAT	24 nt	50.0%	~68°C

Tm values are approximate, calculated at 50 mM NaCl, 0.25 µM oligo using the SantaLucia nearest-neighbor method. Verify exact values with our Tm Calculator.

2. Should You Use TruSeq or Nextera-Style Adapters?

Feature	TruSeq	Nextera / Illumina DNA Prep
Fragmentation	Mechanical (sonication) or enzymatic	Tagmentation (Tn5 transposon)
Adapter attachment	Ligation to dA-tailed fragments	Transposon inserts partial adapters; indexes added by PCR
Adapter structure	Y-shaped (forked) — pre-formed P5/P7 duplex with single-stranded overhang	Mosaic End (ME) sequences — minimal transposon recognition + PCR-added indexes
Input DNA	100 ng – 1 µg	1–500 ng (protocol-dependent)
Hands-on time	~4 hours	~1.5 hours
Insert size range	User-defined (based on fragmentation)	~300 bp (default); adjustable by bead ratio
Sequencing primer sites	TruSeq Read 1/2 primers	Nextera Read 1/2 primers (different sequences)

⚠️ Important: Do Not Mix TruSeq and Nextera Libraries on the Same Lane

TruSeq and Nextera libraries use different sequencing primer binding sites. Pooling them on the same flow cell lane requires loading separate sequencing primers, which is not supported on most Illumina instruments. Always process TruSeq and Nextera libraries on separate lanes or runs unless your core facility has a custom primer loading protocol.

3. Which Dual Indexing Strategy Fits Your Run?

Illumina dual indexing uses two barcode sequences (i7 = Index 1, i5 = Index 2) incorporated into the adapters, enabling multiplexing of dozens to hundreds of samples per sequencing run. The choice between combinatorial and unique dual indexing has significant implications for data quality on modern platforms.

When is combinatorial dual indexing still acceptable?

In CDI, a small set of i5 indexes (e.g., 8) is combined with a set of i7 indexes (e.g., 12) to create 8 × 12 = 96 unique combinations. Individual index sequences are reused across samples — uniqueness comes from the pair. This is cost-effective for small multiplexing on non-patterned flow cell platforms (MiSeq, iSeq 100).

When should you use unique dual indexing?

In UDI, each sample receives a globally unique i5+i7 pair that is never reused within the same pool. If index hopping occurs, the resulting barcode combination matches no valid sample and is automatically filtered as "undetermined" by demultiplexing software (BCL Convert, DRAGEN). This reduces cross-contamination from >1% to <0.01%.

Strategy	Cross-contamination Risk	Platform	Max Samples
Single indexing	High (cannot distinguish hopped reads)	Legacy only	12-24
Combinatorial dual (CDI)	Moderate (~0.1-2% on ExAmp platforms)	MiSeq, iSeq (non-patterned)	96
Unique dual (UDI)	Very low (<0.01%)	All platforms; required for NovaSeq/NextSeq	384 (Set A-D)

4. How Do You Prevent Index Hopping on Illumina Platforms?

Index hopping (also called index switching or barcode swapping) occurs when index sequences from one library molecule become associated with a different library molecule, causing mis-assignment of sequencing reads to the wrong sample. This is a significant concern on patterned flow cell platforms.

Why does hopping happen on patterned flow cells?

On patterned flow cell instruments (NovaSeq 6000/X, NextSeq 1000/2000), Exclusion Amplification (ExAmp) chemistry pre-loads amplification reagents onto the flow cell with library molecules. During cluster generation, free adapter molecules (from incomplete library molecules, degradation, or excess adapters) can hybridize to growing clusters and become incorporated, transferring their index sequence to an unrelated cluster.

Which mitigations matter most?

Strategy	Effectiveness	Implementation
Use UDI indexes	★★★★★	Most effective — invalid pairs filtered during demultiplexing
Clean up free adapters	★★★★	SPRI cleanup (0.8x) before pooling removes adapter dimers and free adapters
Minimize adapters in pool	★★★	Quantify precisely (qPCR/Qubit), avoid over-loading
Fresh pooling	★★	Pool libraries immediately before loading; avoid prolonged storage of pools

5. Which QC Checks Matter Before Ordering Custom Adapters?

Whether using standard Illumina adapters or designing custom ones, thermodynamic validation ensures efficient ligation, proper flow cell binding, and minimal adapter-dimer artifact formation.

Which Tm ranges should adapter regions hit?

Adapter Region	Target Tm	Rationale
Y-adapter duplex region	55-65°C	Stable at room temp for handling; denatures during cluster generation
Flow cell binding (P5/P7)	65-70°C	High Tm ensures stable cluster initiation at 40°C hybridization
Sequencing primer binding	60-65°C	Efficient primer annealing during sequencing-by-synthesis
Index sequences (8 bp)	N/A (too short)	Read as part of sequencing; optimize for color balance (≥2 colors per cycle)

Which structures should force a redesign?

Custom adapter oligos must be evaluated for self-complementarity and hairpin formation. Problematic secondary structures reduce ligation efficiency and can cause adapter-dimers that consume sequencing capacity.

Structure Type	ΔG Threshold	Risk Level	Action
Hairpin	ΔG > −2 kcal/mol	Safe	Proceed
Hairpin	ΔG < −2 kcal/mol	Moderate	Evaluate — may reduce ligation efficiency
Self-dimer	ΔG > −6 kcal/mol	Safe	Proceed
Self-dimer	ΔG < −9 kcal/mol	High	Redesign — will cause adapter-dimer artifacts

Validate adapter oligos using our Secondary Structure Predictor for hairpin and self-dimer analysis, and our Tm Calculator to verify duplex stability at your ligation buffer conditions. For batch validation of custom adapter pools, use Batch Sequence QC.

6. What Changed in Illumina Index Kits in 2025-2026?

Illumina has updated its index product line in 2025:

Rebranding: "IDT for Illumina DNA/RNA UD Indexes" renamed to "Illumina DNA/RNA UD Indexes" — same function, updated branding.
v3 index sequences: Some index sequences updated to v3 for improved color balance during sequencing — ensures ≥2 colors represented per cycle position across the index read.
Transition deadline: Last order date for legacy "IDT for Illumina" indexes was January 30, 2025.
Do not mix: Illumina explicitly states that legacy and new UD index kits should not be combined in the same sequencing run to prevent demultiplexing errors.

The definitive index sequence reference is Illumina's Adapter Sequences Document (#1000000002694), available on the Illumina support page. Always verify your index sequences against this document before library preparation.

7. References

Illumina. (2025). Illumina Adapter Sequences Document. Document #1000000002694.
Illumina. (2024). Index Hopping Filter — Understanding and Mitigating Index Cross-Talk. Illumina Technical Note.
Costello M, et al. (2018). Characterization and remediation of sample index swaps by non-redundant dual indexing on massively parallel sequencing platforms. BMC Genomics, 19:332.doi:10.1186/s12864-018-4703-0
SantaLucia J Jr. (1998). A unified view of polymer, dumbbell, and oligonucleotide DNA nearest-neighbor thermodynamics. PNAS, 95(4):1460-1465.
Adey A, et al. (2010). Rapid, low-input, low-bias construction of shotgun fragment libraries by high-density in vitro transposition. Genome Biology, 11:R119.doi:10.1186/gb-2010-11-12-r119