Back

Variant scoring performance across selection regimes depends on variant-to-gene and gene-to-disease components

Finucane, H. K.; Parsa, S.; Guez, J.; Kanai, M.; Satterstrom, F. K.; Nkambule, L. L.; Daly, M. J.; Seed, C.; Karczewski, K. J.

2024-09-19 bioinformatics
10.1101/2024.09.17.613327 bioRxiv
Show abstract

Variant scoring methods (VSMs) aid in the interpretation of coding mutations and their potential impact on health, but their evaluation in the context of human genetics applications remains inconsistent. Here, we describe GeneticsGym, a systematic approach to evaluating the real-world impact of VSMs on human genetic analysis. We show that the relative performance of VSMs varies across regimes of natural selection, and that both variant-to-gene and gene-to-disease components contribute.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

1
Nature
○ 645 papers in training set
Top 0.4%
23.1%
Similar papers in this journal
2
Nature Communications
○ 5641 papers in training set
Top 7%
19.1%
Similar papers in this journal
3
Nature Genetics
○ 286 papers in training set
Top 0.6%
10.0%
Similar papers in this journal
50% of probability mass above
4
Nature Methods
○ 385 papers in training set
Top 1%
8.2%
Similar papers in this journal
5
Genome Biology
○ 637 papers in training set
Top 2%
5.7%
Similar papers in this journal
6
Nature Biotechnology
○ 172 papers in training set
Top 0.8%
5.0%
7
Cell Genomics
○ 172 papers in training set
Top 2%
2.7%
8
Nucleic Acids Research
◐ 1281 papers in training set
Top 7%
2.2%
9
Science
● 477 papers in training set
Top 5%
1.8%
10
Nature Machine Intelligence
○ 70 papers in training set
Top 2%
1.4%
11
The American Journal of Human Genetics
○ 234 papers in training set
Top 2%
1.1%
12
Nature Biomedical Engineering
○ 47 papers in training set
Top 1%
1.1%
13
Proceedings of the National Academy of Sciences
● 2444 papers in training set
Top 37%
1.1%
14
Bioinformatics
◐ 1204 papers in training set
Top 8%
1.1%
15
Nature Computational Science
○ 55 papers in training set
Top 1%
1.1%
16
eLife
● 5828 papers in training set
Top 61%
1.1%
17
Cell Systems
○ 201 papers in training set
Top 4%
1.1%
18
Nature Structural & Molecular Biology
236 papers in training set
Top 3%
0.9%
19
Science Advances
● 1243 papers in training set
Top 29%
0.9%
20
Genome Research
● 468 papers in training set
Top 6%
0.9%
21
Nature Plants
○ 94 papers in training set
Top 2%
0.9%
22
Communications Biology
○ 993 papers in training set
Top 28%
0.9%
23
Molecular Biology and Evolution
◐ 542 papers in training set
Top 6%
0.5%
24
Molecular Systems Biology
○ 162 papers in training set
Top 4%
0.5%
25
Briefings in Bioinformatics
◐ 354 papers in training set
Top 8%
0.5%
26
Genome Medicine
○ 183 papers in training set
Top 6%
0.5%
27
Nature Chemical Biology
○ 119 papers in training set
Top 4%
0.5%

"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.