Medium to high throughput single cell sequencing: A guide for different scRNAseq applications

07.10.2026

9’

Medium to high throughput single cell sequencing: A guide

Originally published May 9, 2023 | Last updated October 7, 2026

Medium and high throughput single cell RNA sequencing are two approaches to obtain single cell transcriptomic profiles. Medium throughput sequences a small number of cells, from tens to hundreds. High throughput sequences many cells at once, from thousands to millions. The choice between these methodologies should be made with the biological question and experimental design in mind.

Atlases and large databases need efficient, cost-effective analysis of millions of cells. On the other hand, a small, specific population can be missed in high throughput studies, and cell enrichment might be necessary. Moreover, when the cellular input is a limiting factor, it is important to obtain the maximum information from every cell. In these cases, medium throughput is the better choice for generating in-depth transcriptomic profiles.

A third consideration has become as important as cell number: sample number. Perturbation screens, drug panels, and AI virtual cell training sets often require large numbers of samples, and multiplexing becomes an important strategy to streamline the workflow and reduce per-sample costs.

This guide covers both throughput tiers, plus the multiplexing and automation options that allow a high throughput experiment to scale across many samples.

Medium vs high throughput single cell sequencing at a glance

Medium vs high throughput single cell sequencing
	Medium throughput	High throughput
Cells per experiment	Tens to hundreds of cells	Thousands to millions of cells
Representative approaches	SMART-seq3 (plate-based, full-length); AccuraSCOPE (plate-based, DNA and RNA from 96 cells)	Standard single cell and single nucleus RNA-seq; full-length and multi-omic analysis options available
Transcript coverage	Full-length transcripts (SMART-seq3); genome plus transcriptome per cell (AccuraSCOPE)	3′ or 5′ end capture (standard); full-length transcripts (MobiuSCOPE or long-read sequencing)
Reads per cell	> 1 million (full-length)	20,000 – 50,000
Library preparation	Plate-based	Microwell, plate, or droplet-based
Workflow	1 to 2 working days	Hours (microwell/droplet based) to days (plate-based)
Dropout rates	Low	High
Cost per cell	High	Low
Best suited to	Limited input material, rare populations, isoforms and splicing, paired genome and transcriptome	Gene expression, cell type discovery, patient cohorts, atlases, perturbation screens, AIVC training data

Medium throughput: deep data from tens to hundreds of cells

Choose medium throughput when input material is limited, when there is a specific information requirement such as high sensitivity, or when both genomic and transcriptomic information are required on a single cell level.

SMART-seq: full-length transcripts from every cell

Medium throughput methods such as SMART-seq (Hagemann-Jensen et al. 2020) and FLASH-seq (Hahaut et al. 2022) generate full-length cDNA from individual cells, which produces high-quality, deeply sequenced data. The goal is to capture the complete transcript in every cell, so mRNA and isoforms can be quantified precisely.

Library preparation is plate-based. Each cell goes through isolation, lysis, cDNA synthesis, and library preparation, and takes one to two working days from cell lysis to a sequencing-ready library. SMART-seq allows detection of rare transcripts with high sensitivity (Hagemann-Jensen et al. 2020).

Full-length coverage supports detection of alternative splicing events and identification of isoforms and new transcripts with high confidence.

AccuraSCOPE: genome and transcriptome from the same cells

AccuraSCOPE is a plate-based method that amplifies both the DNA and the RNA of 96 individual cells in parallel. Each cell therefore yields a genome and a transcriptome, which links genotype to gene expression in the same cell. This is important for analyzing tumor heterogeneity and drug resistance, where genetic variants may impact disease development and treatment response. The 96-well format runs manually or on an automated liquid handler, and dedicated software handles the data analysis. AccuraSCOPE is suitable when cell numbers are limited and when it is important to map genotype information to the transcriptome at the single cell level.

High throughput: thousands to millions of cells

High throughput methods trade depth per cell for cell number (Ziegenhain et al. 2017). This is the most common approach for profiling cell types, their abundance, and differences between conditions or donors.

Standard single cell and single nucleus RNA-seq

High throughput methods, including Singleron, Drop-seq, and 10x Genomics, capture transcripts with barcoded beads (Macosko et al. 2015, Zheng et al. 2017). Oligos on each bead label every transcript from a single cell. The trade-off is lower coverage per cell in exchange for much higher throughput.

Library preparation is fast because thousands of cells are lysed, their mRNA captured, and libraries built in a single reaction. Reagent volumes per cell are small, which keeps the cost per cell low. Singleron’s workflow takes tissue to a sequencing-ready library within one workday.

Standard single cell sequencing captures most expressed genes but can miss rare transcripts and low-abundance genes.

While standard high throughput single cell sequencing is either 3′ or 5′ biased, it is now possible to obtain full-length or whole-transcriptome sequencing data with modified high throughput protocols such as MobiuSCOPE and RetroSCOPE, or by using long-read sequencing.

Multiplexing: higher sample throughput, lower cost

Multiplexing gives each sample a unique barcode, so several samples can be pooled, processed in one run, and separated computationally afterward (Stoeckius et al. 2018, McGinnis et al. 2019). Pooling lowers the cost per sample and streamlines the processing of large numbers of samples.

Multiplexing for fixed and FFPE samples. RetroSCOPE uses a random priming strategy that captures both poly(A) and non-poly(A) RNA, and is compatible with long-stored FFPE blocks containing degraded and fragmented RNA.

For multiplexing, a sample barcode is added at the reverse transcription (RT) step. Up to 8 samples can be pooled for single nucleus partitioning, cDNA capture, and library preparation.

RetroSCOPE with sample multiplexing is suitable for clinical cohort studies where large batches of archived tissues are analyzed.

Multiplexing for large perturbation and cohort designs. CLindex uses click chemistry to label cell membranes with a unique barcode before pooling (Zhao et al. 2022). The chemical label is species- and cell-type-agnostic.

Up to 96 samples can be pooled using CLindex in a single run. CLindex is compatible with standard automation workflows using liquid handlers.

CLindex is suitable for large-scale single cell experiments, such as drug screening and cell line perturbation studies, and AI virtual cell model training datasets, as discussed below.

Automation: increasing scale and consistency for high throughput single cell sequencing

Tissue dissociation and nucleus extraction can introduce significant variability across samples and operators, and are often bottleneck steps in single cell analysis. This can be addressed through automation, such as the PythoN systems, which process up to 8 samples in parallel following a standardized workflow.

For downstream processing, it is possible to automate cell partitioning, barcoding, reverse transcription, and library preparation in a high throughput workflow. Tensor, an automation instrument for processing up to 32 samples, completes the workflow in 9 hours, and reduces hands-on time by about 90%. Tensor is compatible with a range of single cell analysis, including standard single cell RNA-seq, immune profiling, and full-length single cell RNA-seq.

Combined with multiplexing strategies, automation further increases single cell analysis throughput and enables standardized processing of millions of single cells per run.

The future of high throughput single cell RNA-seq: AI virtual cell

AI virtual cell (AIVC) models are an area where high throughput single cell sequencing, at the scale of millions of cells, is relevant. AIVC models aim to simulate how cells behave across states, tissues, and conditions (Bunne et al. 2024).
To capture the complexity of biology, AIVC models require large quantities of data and adequate coverage of diverse biological contexts (Dibaeinia et al. 2026). The ability to process large numbers of samples across cell lines, animal models, and ultimately clinical samples is key to building the data foundation for AIVC models.

As AIVC models grow, they are likely to drive demand for single cell data at ever larger scales.

Learn more about data generation for AIVC.

How to choose the right throughput

Match the throughput requirement to the question and the sample:

Fewer than 1,000 cells, and you need isoforms or full-length transcripts: Smart-seq-type medium throughput.
A small number of precious cells that need both genome and transcriptome: AccuraSCOPE.
Thousands to millions of cells for cell type discovery, cohorts, or atlases: standard high throughput single cell or single nucleus RNA-seq.
Fixed, FFPE, or archived clinical samples: RetroSCOPE multiplexing.
Many fresh samples or conditions, such as large perturbation experiments: CLindex, with Tensor to automate the run.

References

Bunne C, Roohani Y, Rosen Y, et al. How to build the virtual cell with artificial intelligence: Priorities and opportunities. Cell. 2024;187(25):7045-7063. https://doi.org/10.1016/j.cell.2024.11.015

Dibaeinia P, Babu S, Knudson M, et al. Virtual cells need context, not just scale. bioRxiv. 2026. https://doi.org/10.64898/2026.02.04.703804

Hagemann-Jensen M, Ziegenhain C, Chen P, et al. Single-cell RNA counting at allele and isoform resolution using Smart-seq3. Nat Biotechnol. 2020;38(6):708-714. https://doi.org/10.1038/s41587-020-0497-0

Hahaut, V., Pavlinic, D., Carbone, W. et al. Fast and highly sensitive full-length single-cell RNA sequencing using FLASH-seq. Nat Biotechnol. 2022; 40: 1447-1451. https://doi.org/10.1038/s41587-022-01312-3

Macosko EZ, Basu A, Satija R, et al. Highly parallel genome-wide expression profiling of individual cells using nanoliter droplets. Cell. 2015;161(5):1202-1214. https://doi.org/10.1016/j.cell.2015.05.002

McGinnis CS, Patterson DM, Winkler J, et al. MULTI-seq: sample multiplexing for single-cell RNA sequencing using lipid-tagged indices. Nat Methods. 2019;16(7):619-626. https://doi.org/10.1038/s41592-019-0433-8

Stoeckius M, Zheng S, Houck-Loomis B, et al. Cell Hashing with barcoded antibodies enables multiplexing and doublet detection for single cell genomics. Genome Biol. 2018;19(1):224. https://doi.org/10.1186/s13059-018-1603-1

Zhao X, Sun S, Yu W, et al. Improved ClickTags enable live-cell barcoding for highly multiplexed single cell sequencing. RSC Chem Biol. 2022;3(8):1052-1060. https://doi.org/10.1039/d2cb00046f

Zheng GXY, Terry JM, Belgrader P, et al. Massively parallel digital transcriptional profiling of single cells. Nat Commun. 2017;8:14049. https://doi.org/10.1038/ncomms14049

Ziegenhain C, Vieth B, Parekh S, et al. Comparative analysis of single-cell RNA sequencing methods. Mol Cell. 2017;65(4):631-643.e4. https://doi.org/10.1016/j.molcel.2017.01.023

Planning a single cell project?

Planning a single cell project?

Talk to our scientist about your experimental design.

A post by Yingting Wang

Yingting earned her PhD from the National University of Singapore, specializing in cell biology and tissue engineering. She has eight years of laboratory and commercial experience in single cell multi-omics, including roles in R&D, sales, technical support, and scientific communication.