Illumina Sequencers

2 Illumina sequencers are currently operated by GTF: NovaSeq and MiSeq

NovaSeq 6000

This is the latest sequencer from Illumina. It offers ultimate output for a very competitive price.

4 types of flow cells with different outputs are available.
All flow cells are patterned and ideally require libraries with Unique Dual Indexing.

Flow cells can be used for 100 cycles, 200 cycles, or 300 cycles sequencing runs
These cycles can be split and used either for single end (SR or SE) or paired end runs (PE)
Please discuss actual limitations with GTF staff

Standard settings proposed by GTF are SR 100 cycles / PE 150 cycles / PE 28+90 cycles (10X single cell settings).
The entire flow cell (2 or 4 lanes) must be processed with the same cycling conditions. Requests not matching these standard settings may therefore wait longer (to fill the flow cell) before being processed.

WARNING: Data generated on the NovaSeq may not directly compare to legacy HiSeq data
NovaSeq is a 2-colors system while HiSeq were 4 colors
Novaseq goups Q-scores into 4 bins while HiSeq used > 70 Q-scores (see details)
↠ This change is mostly neutral for gene expression analysis (e.g. RNAseq, bulk or single cell)
↠ It may have a non negligible impact on variant analysis
↠ Analysis pipeline codes may require updating


MiSeq sequencing offers the highest flexibility but is limited by a relatively low throughput. It is therefore recommended for projects that do not require large number of reads, such as targeted sequencing or small genome sequencing. The “nano” option is an inexpensive option for evaluating library quality or pool equimolarity.

  • 1M to 25M reads
  • Single End or Paired End runs
  • Custom read length from 50 nt to 300 nt
MiSeq KitReadlength (Total nt)Million reads*
v2 nano3001
v2 nano5001
v2 micro3004

* Low diversity library sequencing (e.g. 16S or other amplicon based libraries) requires the presence of a stuffer library to artificially increase base content diversity. This reduces the sequencing throughput by 10% to 50%, depending on the spike in level

HiSeq 4000 (discontinued end of 2021)
  • 300 M reads per lane for standard library types
  • Single End or Paired End runs
  • Standard sequencing read length is 150 n

Because of HiSeq4000 index swapping issues (wrong index assigned to some sequencing reads, for details see Illumina website), libraries sequenced on this machine require a unique dual indexing strategy (UDI).

Non-standard read length is available upon demand but must involve a request for an entire flow cell (8 lanes, 2’400M reads).

HiSeq 2500 (discontinued end of 2021)
  • 200 M reads per lane for standard library types
  • Single End or Paired End runs
  • Standard read length is 125 nt for Single End and 100 nt for Paired End

Index swapping is not an issue for sequencing on HiSeq2500. It can be used with any library preparation protocol.

Due to the fact that the sequencing price is equal between HiSeq2500 and HiSeq4000, few requests are submitted for HiSeq2500, resulting in prolonged waiting times. We therefore propose this service either as a full flow cell run (8 lanes) or in a so-called Rapid Mode (see below).

Non-standard read length is available upon request but must involve a request for an entire flow cell (8 lanes).