Back

SPECTER-Based Semantic Triage of Biomedical Literature for Systematic Reviews in Mutational Signature Analysis

Bituin, R. C.; Bokani, A.

2026-07-09 bioinformatics
10.64898/2026.07.06.736558 bioRxiv
Show abstract

Systematic reviews in computational biology require screening large heterogeneous bibliographic sets, especially when topics span computational methods, cancer genomics and statistical modelling. This paper presents a reproducible semantic triage pipeline that combines SPECTER scientific-document embeddings, research-question similarity, proposal-summary similarity and domain keyword coverage to rank candidate studies for systematic review screening. The pipeline was evaluated on 2,231 Covidence records, including 120 final included studies (prevalence = 5.38%), against keyword-only, TF-IDF, BM25, MiniLM, PubMedBERT and SPECTER-only baselines. SPECTER-hybrid achieved the highest average precision (AP = 0.546), recovered 50% of included studies after screening 4.48% of records, and produced an 11.16-fold enrichment over prevalence. Ablation analysis showed that semantic-keyword combinations consistently outperformed single-signal variants. These findings suggest that citation-informed hybrid ranking can support literature triage while retaining human reviewers as final decision-makers.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

1
Briefings in Bioinformatics
354 papers in training set
Top 0.5%
11.1%
2
Bioinformatics
1204 papers in training set
Top 3%
9.8%
3
GigaScience
212 papers in training set
Top 0.2%
7.9%
4
BMC Bioinformatics
457 papers in training set
Top 1%
6.8%
5
Database
61 papers in training set
Top 0.1%
5.5%
6
NAR Genomics and Bioinformatics
242 papers in training set
Top 0.9%
4.3%
7
Nature Communications
5641 papers in training set
Top 32%
4.1%
8
Genome Biology
637 papers in training set
Top 3%
4.1%
50% of probability mass above
9
Nucleic Acids Research
1281 papers in training set
Top 5%
3.5%
10
PLOS ONE
5266 papers in training set
Top 37%
3.3%
11
BioData Mining
22 papers in training set
Top 0.1%
3.3%
12
Research Synthesis Methods
20 papers in training set
Top 0.1%
3.2%
13
Journal of the American Medical Informatics Association
71 papers in training set
Top 1%
2.8%
14
Scientific Reports
3612 papers in training set
Top 47%
2.1%
15
PLOS Biology
486 papers in training set
Top 4%
1.9%
16
Scientific Data
209 papers in training set
Top 1%
1.7%
17
Computational and Structural Biotechnology Journal
242 papers in training set
Top 4%
1.5%
18
Genomics, Proteomics & Bioinformatics
172 papers in training set
Top 1%
1.5%
19
PLOS Computational Biology
1863 papers in training set
Top 15%
1.5%
20
BMC Genomics
406 papers in training set
Top 6%
1.1%
21
Journal of Molecular Biology
232 papers in training set
Top 3%
0.9%
22
PeerJ
308 papers in training set
Top 11%
0.8%
23
Patterns
78 papers in training set
Top 3%
0.8%
24
Bioinformatics Advances
203 papers in training set
Top 4%
0.8%
25
BMC Medical Research Methodology
47 papers in training set
Top 1%
0.8%