Back

Accurate tiling of spatial single-cell data with Tessera

Stein, D. J.; Tran, M.; Korsunsky, I.

2025-01-22 bioinformatics
10.1101/2025.01.17.633630 bioRxiv
Show abstract

Single-cell spatial transcriptomics reveals how cells organize in healthy and diseased tissues. From these data, tissue segmentation analysis defines discrete compartments that organize cells into functional multicellular units. Existing methods for automated tissue segmentation rely on spatial smoothing to define spatially coherent regions but often blur the boundaries between adjacent tissue compartments. We describe Tessera, an algorithm that approaches tissue segmentation through a novel approach, dividing the tissue into small multicellular tiles whose edges track with natural tissue boundaries. Tessera achieves this by incorporating successful tools from edge-preserving smoothing, topological data analysis, and morphology-aware agglomerative spatial clustering. We show that Tessera identifies a range of known anatomical structures, in healthy mouse brain and human lymph nodes, and novel disease-associated niches, in human brain and in lung cancer. Tessera is a general-purpose tool that returns spatially coherent spatial structures with accurate boundaries across a range of spatial transcriptomics and proteomics technologies.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

1
Nature Communications
○ 5641 papers in training set
Top 6%
22.4%
Similar papers in this journal
2
Nature Methods
○ 385 papers in training set
Top 0.4%
18.5%
Similar papers in this journal
3
Genome Biology
○ 637 papers in training set
Top 1%
8.9%
Similar papers in this journal
4
Nature Biotechnology
○ 172 papers in training set
Top 0.2%
8.9%
Similar papers in this journal
50% of probability mass above
5
Nature
○ 645 papers in training set
Top 3%
4.3%
6
Nature Genetics
○ 286 papers in training set
Top 2%
4.3%
Similar papers in this journal
7
Cell Systems
○ 201 papers in training set
Top 2%
2.6%
8
Cell
○ 431 papers in training set
Top 4%
2.4%
9
Science Advances
● 1243 papers in training set
Top 17%
2.1%
10
eLife
● 5828 papers in training set
Top 48%
1.9%
11
Science
● 477 papers in training set
Top 5%
1.7%
12
Communications Biology
○ 993 papers in training set
Top 14%
1.7%
13
Nucleic Acids Research
◐ 1281 papers in training set
Top 10%
1.4%
14
Bioinformatics
◐ 1204 papers in training set
Top 7%
1.3%
15
Cell Genomics
○ 172 papers in training set
Top 3%
1.1%
16
Cell Reports Methods
○ 165 papers in training set
Top 2%
1.1%
17
Cell Reports
○ 1498 papers in training set
Top 23%
1.1%
18
Nature Biomedical Engineering
○ 47 papers in training set
Top 1%
1.1%
19
Nature Cell Biology
○ 118 papers in training set
Top 3%
0.8%
20
Molecular Systems Biology
○ 162 papers in training set
Top 3%
0.8%
21
Proceedings of the National Academy of Sciences
● 2444 papers in training set
Top 41%
0.8%
22
Nature Computational Science
○ 55 papers in training set
Top 2%
0.6%
23
iScience
○ 1154 papers in training set
Top 41%
0.6%
24
Advanced Science
○ 286 papers in training set
Top 11%
0.6%

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.