Human tissue comprises dozens of spatially organised cell types, all working together to maintain normal tissue function. Similarly, diseased tissue also regulates its spatial organisation, and the location of individual cell types in the tissues contributes to the disease development. For example, in cancer patients, tumour tissues with variable spatial organisation can lead to significantly different treatment responses and patient outcomes.
Spatial transcriptomics is a transformative experimental technology that now enables researchers to spatially profile tissue samples, revealing the spatial organisation of individual cells within the tissue.
Spatial transcriptomics simultaneously measures the activity of approximately 25,000 genes in all cells, providing insights into how cells in different locations regulate themselves. With spatial transcriptomics, researchers can now profile hundreds of thousands of cells in human specimens and determine the spatial location and transcriptome of each cell.
Although the spatial transcriptomics methods are powerful, their downstream data analyses are currently complicated and computationally intensive. Without easy-to-use data analytical tools, the true potential of the spatial data for biological discovery remains limited.
To overcome this challenge, a team led by Senior Researcher Mitro Miihkinen from the Institute for Molecular Medicine Finland (FIMM) at the University of Helsinki, has created a computational platform that simplifies this complex work.
The team, including researchers from FIMM and the iCAN Digital Precision Cancer Medicine flagship (part of the Research Council of Finland’s flagship programme), developed a marker-based cell typing method scType for spatial transcriptomics data.
Importantly, the scType method is both accurate and ultra-fast, allowing analysis to be performed on a laptop computer in a matter of seconds, as their recent publication in the journal Bioinformatics demonstrates.
“Previously developed cell typing methods for spatial transcriptomics have required a lot of computational power and reference data. With the significant improvement in spatial resolution of these assays, we hypothesised that this should allow the development of computationally efficient marker-based cell typing tools,” explains Doctoral Researcher Kristen Nader from FIMM, the first author of the study and the main developer of the method.
The team demonstrated the feasibility of the scType method by re-analysing different spatial samples generated by various spatial transcriptomics assays. The results showed that the scType platform was extremely accurate in assigning cell type labels to different spatial locations while also being highly memory-efficient.
“ScType will accelerate unbiased spatial profiling of tissues by being extremely easy and versatile to use. Its speed is also unparalleled, especially now as dataset sizes are getting larger and larger, which places more strain on computing environments,” says Mitro Miihkinen, the senior author of the work.
To promote its wide application, either as a stand-alone tool or in conjunction with other popular transcriptomics data analysis software, the group has deployed scType for spatial transcriptomics both as an open-source R and python packages.
Original publication: Nader K, Tasci M, Ianevski A, Erickson A, Verschuren EW, Aittokallio T, Miihkinen M. ScType enables fast and accurate cell type identification from spatial transcriptomics data. Bioinformatics 2024 Jul 1;40(7):btae426
Further information:
Mitro Miihkinen, PhD, Senior Researcher
Institute for Molecular Medicine Finland FIMM, HiLIFE, University of Helsinki
Email: mitro.miihkinen@helsinki.fi
Kristen Nader, Doctoral Researcher
Institute for Molecular Medicine Finland FIMM, HiLIFE, University of Helsinki
Email: kristen.nader@helsinki.fi