Back

Cell type-specific signal analysis in EWAS

Breeze, C.

2021-05-23 genomics
10.1101/2021.05.21.445209 bioRxiv
Show abstract

Hundreds of epigenome-wide association studies (EWAS) have been performed, successfully identifying replicated epigenomic signals in processes such as ageing and smoking. Despite this progress, it remains a major challenge in EWAS to detect both cell type-specific and cell type confounding effects impacting study results. One way to identify these effects is through eFORGE (experimentally derived Functional element Overlap analysis of ReGions from EWAS), a published tool that uses 815 datasets from large-scale mapping studies to detect enriched tissues, cell types and genomic regions. Here, I show that eFORGE analysis can be extended to EWAS differentially variable positions (DVPs), identifying target cell types and tissues. In addition, I also show that eFORGE tissue-specific enrichment can be detected for sites below EWAS significance threshold. I develop on these and other analysis examples, extending our knowledge of eFORGE cell type- and tissue-specific enrichment results for different EWAS.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

1
BMC Genomics
○ 406 papers in training set
Top 0.1%
14.9%
Similar papers in this journal
2
BMC Bioinformatics
○ 457 papers in training set
Top 0.3%
14.9%
Similar papers in this journal
3
Genes
○ 144 papers in training set
Top 0.3%
5.4%
Similar papers in this journal
4
Frontiers in Genetics
○ 230 papers in training set
Top 0.5%
5.4%
Similar papers in this journal
5
NAR Genomics and Bioinformatics
◐ 242 papers in training set
Top 1.0%
4.3%
Similar papers in this journal
6
Computational and Structural Biotechnology Journal
● 242 papers in training set
Top 2%
3.2%
7
Epigenetics
○ 50 papers in training set
Top 0.2%
3.2%
50% of probability mass above
8
PLOS ONE
● 5266 papers in training set
Top 37%
3.2%
9
Bioinformatics Advances
◐ 203 papers in training set
Top 2%
3.1%
10
Epigenetics & Chromatin
52 papers in training set
Top 0.2%
3.0%
11
Bioinformatics
◐ 1204 papers in training set
Top 6%
2.7%
12
International Journal of Molecular Sciences
○ 494 papers in training set
Top 5%
2.4%
13
Scientific Reports
○ 3612 papers in training set
Top 41%
2.4%
14
BMC Medical Genomics
○ 50 papers in training set
Top 0.3%
2.4%
15
BMC Genomic Data
○ 13 papers in training set
Top 0.1%
2.1%
16
Genome Biology
○ 637 papers in training set
Top 5%
2.1%
17
Briefings in Bioinformatics
◐ 354 papers in training set
Top 4%
2.1%
18
Nucleic Acids Research
◐ 1281 papers in training set
Top 8%
2.1%
19
Epigenomics
○ 11 papers in training set
Top 0.1%
1.7%
20
Genomics
○ 64 papers in training set
Top 0.8%
1.7%
21
GigaScience
◐ 212 papers in training set
Top 3%
1.4%
22
Genome Medicine
○ 183 papers in training set
Top 4%
1.1%
23
Methods
○ 34 papers in training set
Top 0.5%
1.0%
24
Genetic Epidemiology
○ 55 papers in training set
Top 0.7%
0.8%
25
Mobile DNA
○ 31 papers in training set
Top 0.3%
0.8%
26
PLOS Computational Biology
● 1863 papers in training set
Top 22%
0.6%
27
Human Genomics
○ 21 papers in training set
Top 0.6%
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.