Smart-seq3

Profiling of gene expression levels at single-cell resolution.

Protocol for sequencing of full-length single-cell transcripts, based on the method described in Hagemann-Jensen et al. in Nature Biotechnology 2020. Detailed protocol is available at protocols.io.

Pros to this method: No limit on cell size, no need to keep cells fresh for transport. Full-length transcript information. Cheaper than droplet-based methods for small sample sizes.

Cons to this method: Limited number of cells, needs access to FACS sorting facility. Labour-intensive sample preparation.

NOTE! Smart-seq3 projects are ordered through the KI iLabs system (search for ESCG).

Sample Requirements

Single cells, sorted into 384-well plates (provided by us) containing lysis buffer and ERCC RNA-spike in. Please see our full sample requirements for further information about providing NGI with samples for SS3 sequencing.

Lysis plates

Lysis plates are made as needed (typically once per month), please contact us so we can reserve the required number of sample & validation plates for your project.

Empty lysis plates should be stored at -20°C.

Plates with lysed cells should be stored at -80°C and returned (on dry ice) to NGI.

How to evaluate the sample quality

In addition to the sample plates, we also make validation plates, with lysis buffer in only the first two columns.

To ensure high quality samples and optimal FACS sorting, we recommend a minimum of 1 validation plate per cell type or experimental batch. When the plates are received at NGI, we will process the validation plate(s) first, and the amount and quality of cDNA is assessed before proceeding with full sample plates.

What we do with your samples

Reverse transcription and preamplification PCR, followed by cDNA purification, quantification & QC.

Library Preparation & QC

Tagmentation & clean-up, size selection. Library QC & pooling.

Sequencing

The Smart-seq3 libraries are sequenced on an Illumina sequencer, eg. NovaSeq XPlus, with a default sequencing depth of 150k reads/sample (or multiples thereof) and an 85-10-10-133 read setup.

Bioinformatics

BCL files are converted to fastq-files using bcl2fastq.

The zUMIs pipeline (Parekh et al. Gigascience, 2018) is run to process UMIs and internal reads.

Last Updated: 1st March 2024