Back

The Impact of Variant Calling on Substitution Mutational Signature Inference

Jiang, Z.; Au, J.; Kazachkova, M.; Diaz-Gay, M.; Vangara, R.; Alexandrov, L. B.

2025-12-12 bioinformatics
10.64898/2025.12.09.693327 bioRxiv
Show abstract

Identifying mutational signatures is a key component of cancer genomics studies, yet the influence of variant calling strategies on signature extraction has not been systematically evaluated. Here, we analyzed over 8,900 whole exomes from The Cancer Genome Atlas (TCGA) and over 1,800 whole genomes from the Pan-Cancer Analysis of Whole Genomes (PCAWG) consortium to assess how mutation callers shape de novo single-base substitution (SBS) signatures. We found that consensus calling yielded stable de novo signatures across reference genomes and pipeline versions, whereas individual callers introduced false-positive SBSs that manifested as artifactual signatures that were reproducibly detected by three independent signatures extraction tools. A minimal consensus approach requiring agreement between only two variant calling algorithms effectively removed these artifacts while preserving true biological signal. Together, these results establish consensus variant calling as essential for robust inference of de novo SBS mutational signatures and provide practical guidelines for distinguishing genuine mutational processes from technical artifacts.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

1
Nature Communications
○ 5641 papers in training set
Top 6%
22.7%
Similar papers in this journal
2
Cell Systems
○ 201 papers in training set
Top 0.4%
7.4%
Similar papers in this journal
3
Nature Biotechnology
○ 172 papers in training set
Top 0.4%
6.8%
Similar papers in this journal
4
Nature Methods
○ 385 papers in training set
Top 2%
5.6%
Similar papers in this journal
5
Nucleic Acids Research
◐ 1281 papers in training set
Top 4%
4.4%
Similar papers in this journal
6
Genome Biology
○ 637 papers in training set
Top 3%
4.4%
50% of probability mass above
7
Nature Genetics
○ 286 papers in training set
Top 2%
4.4%
8
Genome Medicine
○ 183 papers in training set
Top 0.9%
4.1%
9
The American Journal of Human Genetics
○ 234 papers in training set
Top 1%
3.6%
10
Cell Genomics
○ 172 papers in training set
Top 2%
2.4%
11
Communications Biology
○ 993 papers in training set
Top 10%
2.2%
12
Briefings in Bioinformatics
◐ 354 papers in training set
Top 4%
2.0%
13
Cell Reports Medicine
○ 153 papers in training set
Top 2%
1.8%
14
Scientific Reports
○ 3612 papers in training set
Top 51%
1.8%
15
Genome Research
● 468 papers in training set
Top 4%
1.5%
16
Science Advances
● 1243 papers in training set
Top 22%
1.5%
17
npj Precision Oncology
○ 53 papers in training set
Top 1%
1.4%
18
Cancer Research
● 130 papers in training set
Top 2%
1.4%
19
iScience
○ 1154 papers in training set
Top 23%
1.4%
20
Cell Reports
○ 1498 papers in training set
Top 22%
1.1%
21
NAR Genomics and Bioinformatics
◐ 242 papers in training set
Top 3%
1.1%
22
Bioinformatics
◐ 1204 papers in training set
Top 8%
1.1%
23
PLOS Computational Biology
● 1863 papers in training set
Top 18%
1.0%
24
Proceedings of the National Academy of Sciences
● 2444 papers in training set
Top 40%
0.9%
25
Nature Biomedical Engineering
○ 47 papers in training set
Top 2%
0.9%
26
eLife
● 5828 papers in training set
Top 69%
0.6%
27
Cell Reports Methods
○ 165 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.