Available
Illumina index and barcode collision checker
A spreadsheet can list barcodes. It rarely computes Hamming distance, reverse complements, or whether i7/i5 reuse will fail unique-dual requirements—until demultiplexing assigns reads to the wrong sample.
For core-lab staff and library-prep leads who need a pass/fail collision report before the pool goes on the sequencer, not after the run.
How it works
01
Provide indexes
Upload a CSV or XLSX, paste a list, or drop the Data section of an Illumina sample sheet. Use sample_id, index (i7), and optional index2 (i5). Paste oligo bases, not catalog IDs such as D701.
02
Set the rules
Choose the minimum Hamming distance (3 is the usual Illumina default), whether reverse complements count, whether unique dual indexes are required, and an optional index-read truncation.
03
Review findings
See duplicates, pairs below the distance threshold, reverse-complement collisions, and combinatorial i7/i5 reuse. Unsatisfied rules are listed explicitly.
04
Export the record
Download the finding table, the sample table, and a JSON record of the inputs, settings, and software version.
Constraints and checks
- Single-index and dual-index pools
- Configurable minimum Hamming distance (default 3)
- Exact duplicates on i7, i5, or the dual pair
- Reverse-complement collisions from mixed i5 orientations
- Unique dual versus combinatorial reuse (index-hopping risk)
- Optional truncation to the sequenced index-read length
- CSV/XLSX, pasted lists, and Illumina sample-sheet Data sections
- Catalog IDs such as D701 labeled as names, not sequences
What it will not do
- It does not look up catalog names such as D701 to a sequence.
- It does not rewrite a sample sheet or convert i5 orientation for a named instrument.
- It does not simulate demultiplexing yield or index-hopping rates.
- It does not design a new barcode set.
Why distance 1 fails a pool
- Hamming distance for Illumina indexes
Demultiplexing needs leftover substitutions after an error. Distance 1 cannot provide that, and combinatorial dual indexes add an index-hopping failure even when the pair looks unique.
Trust, files, and price
- Free, including export
- No account required
- Runs in your browser; index tables never leave your machine
- Deterministic report for a given table, settings, and version
- No patient-identifiable data by design
Free, including CSV export and the JSON record. No account.
Get product updates
No account is created. We will only use this address for occasional product notes about this tool.
Questions
- What Hamming distance should I use?
- 3 is the usual Illumina recommendation for a pool: one substitution of error still leaves the read closer to the true index than to any other. Distance 1 or 2 in the same lane is a failed pool for most chemistries. You can lower or raise the threshold; the report records whatever you set.
- What is a unique dual index?
- Each sample’s i7 is unique in the pool, and each i5 is unique. Combinatorial dual indexing reuses i7 or i5 across samples. On patterned flow cells that reuse is an index-hopping risk even when the i7+i5 pair is unique. The checker flags that reuse when “Require unique dual indexes” is on.
- Can I paste D701 or UDP0001?
- No. Those are catalog names. The checker needs the oligo bases (A/C/G/T) as they will appear in the sample sheet. Catalog IDs are labeled “name, not sequence” and fail the pool; they are not looked up.
- Do you convert i5 for NovaSeq versus MiSeq?
- No. Paste the sequences as they will appear in the sample sheet. Reverse-complement checking is for catching a mixed-orientation list, not for rewriting oligo catalogs.
- Is this free?
- Yes. The checker runs in your browser and export is included. Index tables are not uploaded.
- What file do I get back?
- A CSV of pairwise findings, a CSV of per-sample status, and a JSON record of the version, settings, and inputs that produced the report.