Back

Advances in mixed cell deconvolution enable quantification of cell types in spatially-resolved gene expression data

Danaher, P.; Kim, Y.; Nelson, B.; Griswold, M.; Yang, Z.; Piazza, E.; Beechem, J. M.

2020-09-29 bioinformatics
10.1101/2020.08.04.235168 bioRxiv
Show abstract

We introduce SpatialDecon, an algorithm for quantifying cell populations defined by single cell RNA sequencing within the regions of spatially-resolved gene expression studies. It obtains cell abundance estimates that are spatially-resolved, granular, and paired with highly multiplexed gene expression data. SpatialDecon incorporates several advancements in the field of gene expression deconvolution. We propose an algorithm based in log-normal regression, attaining sometimes dramatic performance improvements over classical least-squares methods. We compile cell profile matrices for 27 tissue types. We identify genes whose minimal expression by cancer cells makes them suitable for immune deconvolution in tumors. And we provide a lung tumor dataset for benchmarking immune deconvolution methods. In a lung tumor GeoMx DSP experiment, we observe a spatially heterogeneous immune response in intricate detail and identify 7 distinct phenotypes of the localized immune response. We then demonstrate how cell abundance estimates give crucial context for interpreting gene expression results.

Matching journals

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

1
Nature Methods
○ 385 papers in training set
Top 0.7%
13.1%
Similar papers in this journal
2
Nature Communications
○ 5641 papers in training set
Top 14%
12.8%
Similar papers in this journal
3
Nature Biotechnology
○ 172 papers in training set
Top 0.1%
11.2%
Similar papers in this journal
4
Genome Biology
○ 637 papers in training set
Top 1%
8.0%
Similar papers in this journal
5
Cell Reports Methods
○ 165 papers in training set
Top 0.1%
7.4%
Similar papers in this journal
50% of probability mass above
6
Bioinformatics
◐ 1204 papers in training set
Top 4%
6.3%
7
Nucleic Acids Research
◐ 1281 papers in training set
Top 5%
4.1%
8
Cell Systems
○ 201 papers in training set
Top 1%
3.3%
9
eLife
● 5828 papers in training set
Top 41%
2.5%
10
Cancer Research Communications
● 51 papers in training set
Top 0.5%
2.4%
11
Molecular Systems Biology
○ 162 papers in training set
Top 1%
2.1%
12
Briefings in Bioinformatics
◐ 354 papers in training set
Top 4%
2.1%
13
Genome Medicine
○ 183 papers in training set
Top 2%
2.1%
14
Communications Biology
○ 993 papers in training set
Top 12%
1.9%
15
PLOS Computational Biology
● 1863 papers in training set
Top 14%
1.7%
16
Scientific Reports
○ 3612 papers in training set
Top 56%
1.5%
17
Cancer Research
● 130 papers in training set
Top 2%
1.4%
18
Science Advances
● 1243 papers in training set
Top 25%
1.1%
19
Genome Research
● 468 papers in training set
Top 6%
1.0%
20
Patterns
○ 78 papers in training set
Top 3%
0.9%
21
iScience
○ 1154 papers in training set
Top 35%
0.9%
22
NAR Genomics and Bioinformatics
◐ 242 papers in training set
Top 5%
0.6%
23
npj Systems Biology and Applications
○ 125 papers in training set
Top 2%
0.6%
24
BMC Bioinformatics
○ 457 papers in training set
Top 6%
0.6%
25
Nature Genetics
○ 286 papers in training set
Top 5%
0.6%
26
Molecular Biology of the Cell
● 311 papers in training set
Top 4%
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.