Back

Generation of Super-resolution Images from Barcode-based Spatial Transcriptomics Using Deep Image Prior

Park, J.; Cook, S.; Lee, D.; Choi, J.; Yoo, S.; Im, H.-J.; Lee, D.; Choi, H.

2023-06-28 bioinformatics
10.1101/2023.06.26.546529 bioRxiv
Show abstract

Spatial transcriptomics (ST) has revolutionized the field of biology by providing a powerful tool for analyzing gene expression in situ. However, current ST methods, particularly barcode-based methods, have limitations in reconstructing high-resolution images from barcodes sparsely distributed in slides. Here, we present SuperST, a novel algorithm that enables the reconstruction of dense matrices from low-resolution ST libraries. SuperST based on deep image prior reconstructs spatial gene expression patterns as image matrices. SuperST allows gene expression mapping to better reflect immunofluorescence (IF) images. Compared with previous methods, SuperST generated output images that more closely resembled IF images for given gene expression maps. Additionally, SuperST overcomes the limitations inherent in IF images, highlighting its potential applications in the realm of spatial biology. By providing a more detailed understanding of gene expression in situ, SuperST has the potential to contribute to comprehensively understanding biology from various tissues.

Matching journals

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

1
Briefings in Bioinformatics
◐ 354 papers in training set
Top 0.1%
18.9%
Similar papers in this journal
2
Scientific Reports
○ 3612 papers in training set
Top 2%
12.2%
Similar papers in this journal
3
Bioinformatics
◐ 1204 papers in training set
Top 3%
10.0%
Similar papers in this journal
4
Computational and Structural Biotechnology Journal
● 242 papers in training set
Top 0.3%
6.9%
Similar papers in this journal
5
BMC Bioinformatics
○ 457 papers in training set
Top 2%
5.6%
Similar papers in this journal
50% of probability mass above
6
GigaScience
◐ 212 papers in training set
Top 0.6%
5.0%
7
PLOS Computational Biology
● 1863 papers in training set
Top 11%
2.8%
8
PLOS ONE
● 5266 papers in training set
Top 40%
2.5%
9
Communications Biology
○ 993 papers in training set
Top 10%
2.2%
10
npj Systems Biology and Applications
○ 125 papers in training set
Top 0.9%
1.8%
11
iScience
○ 1154 papers in training set
Top 18%
1.7%
12
Frontiers in Bioinformatics
○ 49 papers in training set
Top 0.6%
1.5%
13
Scientific Data
○ 209 papers in training set
Top 2%
1.4%
14
Computers in Biology and Medicine
○ 128 papers in training set
Top 3%
1.4%
15
Genomics, Proteomics & Bioinformatics
188 papers in training set
Top 1%
1.2%
16
Bioinformatics Advances
◐ 203 papers in training set
Top 4%
1.2%
17
Nature Communications
○ 5641 papers in training set
Top 49%
1.2%
18
Biological Imaging
◐ 15 papers in training set
Top 0.2%
1.2%
19
Journal of Pathology Informatics
○ 15 papers in training set
Top 0.2%
0.9%
20
Nucleic Acids Research
◐ 1281 papers in training set
Top 13%
0.9%
21
IEEE Journal of Biomedical and Health Informatics
● 37 papers in training set
Top 1%
0.9%
22
Patterns
○ 78 papers in training set
Top 2%
0.9%
23
Advanced Science
○ 286 papers in training set
Top 10%
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.