Back

Preserving tissue structure through density-based spatial analysis with scider

Li, M.; Liu, N.; Nguyen, Q. H.; Chen, Y.

2025-09-17 bioinformatics
10.1101/2025.09.11.675745 bioRxiv
Show abstract

In spatial transcriptomics, most existing approaches for spatial domain detection analyze slide-wide gene expression patterns while integrating information from local neighborhoods or nearest neighbors. However, such global strategies can overlook tissue organization and obscure regional heterogeneity. To address this, we present scider, a framework for spatial data analysis to preserve tissue structures. scider defines spatial domains from local cellular organization and distribution, instead of relying on global clustering of gene expression and neighborhood profiles. Kernel density estimation (KDE) is used to characterize the spatial distribution of cells, enabling the robust and unbiased identification of regions of interest (ROIs) beyond histological annotations. We show that ROIs make multi-sample analysis possible and reveal spatial patterns that are undetectable in single-sample analysis. Additionally, KDE-derived contour lines define regions of similar cell density, supporting cell type composition and differential expression analyses along continuous spatial gradients.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

1
Nature Communications
○ 5641 papers in training set
Top 8%
18.7%
Similar papers in this journal
2
Genome Biology
○ 637 papers in training set
Top 0.5%
15.2%
Similar papers in this journal
3
Nature Methods
○ 385 papers in training set
Top 2%
6.8%
Similar papers in this journal
4
Genome Medicine
○ 183 papers in training set
Top 0.7%
5.2%
Similar papers in this journal
5
Communications Biology
○ 993 papers in training set
Top 2%
4.9%
Similar papers in this journal
50% of probability mass above
6
Cell Systems
○ 201 papers in training set
Top 1.0%
4.4%
7
Cell Reports Methods
○ 165 papers in training set
Top 0.4%
4.1%
8
Nucleic Acids Research
◐ 1281 papers in training set
Top 5%
3.2%
9
Advanced Science
○ 286 papers in training set
Top 2%
3.2%
10
Bioinformatics
◐ 1204 papers in training set
Top 6%
2.5%
11
Briefings in Bioinformatics
◐ 354 papers in training set
Top 4%
2.4%
12
Nature Biotechnology
○ 172 papers in training set
Top 2%
2.4%
13
PLOS Computational Biology
● 1863 papers in training set
Top 13%
2.1%
14
iScience
○ 1154 papers in training set
Top 13%
2.1%
15
eLife
● 5828 papers in training set
Top 48%
1.9%
16
Science Advances
● 1243 papers in training set
Top 20%
1.7%
17
GigaScience
◐ 212 papers in training set
Top 3%
1.4%
18
Cancer Research
● 130 papers in training set
Top 2%
1.1%
19
Scientific Reports
○ 3612 papers in training set
Top 62%
1.1%
20
Genomics, Proteomics & Bioinformatics
188 papers in training set
Top 2%
1.1%
21
Small Methods
○ 29 papers in training set
Top 0.5%
1.1%
22
Proceedings of the National Academy of Sciences
● 2444 papers in training set
Top 40%
0.9%
23
Molecular Systems Biology
○ 162 papers in training set
Top 3%
0.9%
24
Cell Reports
○ 1498 papers in training set
Top 26%
0.8%
25
PLOS ONE
● 5266 papers in training set
Top 61%
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.