Sequencing by Illumina

Sequencing by Illumina instruments is provided by NGI Stockholm and NGI Uppsala (the SNP&SEQ Technology platform).

Illumina sequencing generate by far the largest amount of data per run. This makes Illumina sequencing ideal for projects with a large number of samples and/or whole genome re-sequencing (WGS) of larger genomes (>200 Mb), transcriptome sequencing, including small RNA (miRNA) sequencing. It is also suitable for whole exome sequencing, targeted re-sequencing, metagenomics studies, amplicon sequencing, de novo sequencing, methylation analysis, chromatin-immunoprecipitated (ChIP) sequencing and more.

The following Illumina instruments are available:

  • MiSeq v2: Gives >10 M reads per run. Read lengths up to 2x250bp
  • MiSeq v3: Gives >18 M reads per run. Read lengths up to 2x300bp
  • HiSeq 2500 High Output mode: Gives >190 M reads per lane. Read lengths up to 2x125bp.
  • HiSeq 2500 Rapid Run mode: Gives >230 M reads per flowcell. Read lengths up to 2x250bp.
  • HiSeq X: Gives approximately 350 M read-pairs per lane. Read length 2x150bp. Only for WGS at 30X coverage or more.

NGI Stockholm and NGI Uppsala offer both library preparation and sequencing. The decision on which library preparation kit, instrument and set up are best for your particular project depend on various factors such as research question, amount and quality of starting material, requested amount of data and so on. For any inquiries on what is best for your project, please contact the project coordinators at NGI Stockholm and NGI Uppsala (the SNP&SEQ Technology platform) (see contact information) or request a meeting.

Pooling threshold - NGI Stockholm

When pooling samples in a lane we aim, by default to an even distribution of all the samples. In other words, when loading n samples in a lane we aim to produce the same amount of data for each sample. In certain cases user might be interested in pooling samples in the same lane aiming to different percentages, this can be discussed with project coordinators.

However, it is not always possible to distribute the libraries in the pool completely equally due to factors such as measuring error in library quantitation, pipetting error etc. At NGI Stockholm we guarantee a minimum amount of reads to be produced when pooling samples:

minimum_reads = 75% * (expected_lane_yield / number_of_samples)

This threshold is not applied for pools done by the users.