Kenny Roberts, Bayraktar lab, Wellcome Sanger Institute

High Throughput Spatial Genomics

Sanger-EBI Initative for Large-Scale Spatial Genomics

Spatial genomics, a new frontier in molecular biology, aims to assay the genomic information of single cells within their native tissue environment. This collaborative initiative between multiple teams and departments at the Wellcome Sanger Institute, including Cellular Genetics, CASM, and Scientific Operations, as well as EMBL-EBI, combines high-throughput sequencing and microscopy to generate spatial genomic data at scale. We aim to harness these technologies towards building 3D cellular atlases of human tissues and analysing hundreds of patient tissue samples.

Mission Statement

Recent technological advances in spatially resolved RNA-sequencing and high-multiplexed RNA imaging have pioneered the spatial transcriptomic analysis of tissues. The promises of these exciting spatial transcriptomics technologies – 3D cellular atlases of entire tissues/organisms and spatial biology of hundreds of patient samples – can only be realized with large-scale experimental and computational platforms. 

The High-Throughput Spatial Genomics (HTSG) initiative utilises cutting-edge technologies to generate spatial omics data at scale. We combine single cell and spatial RNA-sequencing, high-throughput microscopy and quantitative image data analysis to enable large-scale spatial genomic studies. Our activities strongly tie into the Human Cell Atlas as well as genomic technology development and cancer biology.

Spatial Technologies

The Initiative employs diverse sequencing and imaging technologies for spatial transcriptomic data generation.

Mapping cell types in spatial RNA-seq data using cell2location

Visium Spatial Transcriptomics: 

Visium technology (10X Genomics) enables spatial RNA-sequencing (RNA-seq) by positionally capturing RNAs from thin tissue sections onto an oligonucleotide array. The Visium arrays provide a spatial resolution of 55 microns – capturing multiple cells at each tissue position. To computationally extend the resolution of the Visium assay and other spatial RNA-seq methods to map individual cell types, we developed the cell2location model. Cell2location integrates single cell and spatial transcriptomic data to comprehensively map cell types across complex tissues.

We utilise both Visium for fresh frozen samples, and the probe hybridisation-sequencing hybrid Visium CytAssist for FFPE tissues.


ISS in the human brain. We image entire large tissue sections (top left panel) at high resolution (bottom panels) to quantify single cell gene expression in situ.

In Situ Sequencing and Xenium:

We use both manual in situ sequencing (CARTANA) and automated 10x Genomics Xenium to quantify gene expression at single-cell resolution in tissues. In situ sequencing (ISS) provides a cyclic and combinatorial barcoded FISH approach to label up to 500 different transcripts in a multiplexed manner. We have coupled ISS with fast tissue imaging and bespoke image processing pipelines to generate a high-throughput pipeline for various human tissues.











Hiplex RNAscope:

This protocol provides highly-sensitive measurements of small gene panels in tissues. Hiplex RNAscope provides a cyclic and signal amplified FISH approach to label up to 12 transcripts in a multiplexed manner.


Hiplex RNAscope labelling of 12 transcripts in the developing human hindlimb.


Hiplex RNAscope labelling of neuronal subtype markers in the mouse cerebral cortex.



Perkin Elmer Opera Phenix HCS

Phenix multi-camera SD microscope provides fast imaging of large tissue sections.

We have adapted this High Content Screening system for automated imaging of tissue sections labeled with ISS or RNAscope smFISH. To provide high-throughput imaging at high spatial resolution, this system was assembled with a Yokogawa spinning disk and 4 cameras as well as automated water immersion objectives and software autofocus.


Spinning Disk Confocal Microscope
“Nemo” custom spinning Disk Confocal Microscope

CAIRN “Nemo” Spinning Disk

The most recent addition to the facility, “Nemo” is a custom dual-camera spinning disk confocal microscope assembled around a CrestOptics X-light V3 by Cairn Research. The completely open hardware and software the microscope runs on gives the facility more flexibility to implement automated cyclic imaging protocols for ISS and other high-multiplexed RNA imaging methods.

CAIRN “Ahab” (coming soon)

The next planned instrument for the facility will be a larger, four-camera version of “Nemo” also to be assembled and installed by Cairn Research


Image Data Infrastructure

Cyclic Microscopy Image Analysis Pipelines

Cell segmentation of organoid images by cellpose, visualised on napari.


Our computational pipelines and IT platforms enable processing ISS & RNAScope image datasets at scale. Our bespoke workflows enable registration of large cyclic images to sub pixel accuracy, robust barcode decoding and cellular segmentation across a variety of different organs and organoids. We use python based tools cell segmentation (e.g. cellpose) and visualisation (napari). To enable large-scale application, we deploy nextflow/jupyter based pipelines on high performance GPUs. These IT activities are supported by the Cellgen IT and Scientific IT teams at Sanger.



Browsing a high resolution human lung image dataset on OMERO.

Image data management and visualisation via OMERO

All image data produced by the facility is stored in our OMERO databases maintained by the Cellgen IT team. Our Image Data Resource (IDR) instances allow us to visualise and browse large tissue image datasets in the “Google Maps” format and also enable us to share image datasets with external collaborators. These databases are maintained on petabyte-scale high performance storage systems.


BioImage Archive

We collaborate with the new BioImage Archive to disseminate our large-scale spatial transcriptomic image datasets to the researcher community.


Core team

Photo of Dr Kwasi Amoako Kwakwa

Dr Kwasi Amoako Kwakwa

Visiting Scientist

Photo of Dr Tong LI

Dr Tong LI

Senior Software Developer

Photo of Dr Fani Memi

Dr Fani Memi

Senior Staff Scientist

Photo of Dr Kenny Roberts

Dr Kenny Roberts

Senior Staff Scientist

Photo of Dr Benjamin (Ben) John Woodhams

Dr Benjamin (Ben) John Woodhams

Postdoctoral Fellow in High Throughput Spatial Genomics

Previous team members

Photo of Zoi Olvia  Katsirea

Zoi Olvia Katsirea

Advanced Research Assistant

Photo of Dr Jun Sung Park

Dr Jun Sung Park

Visiting Scientist

Photo of Aleksandra (Ola) Tarkowska

Aleksandra (Ola) Tarkowska

Solutions Architect


We work with the following groups