Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 58 of the 58 references it cites were found in a registry or an index.

Safetyome and specialized panels for over 3,000 phenotypes: a systematic and translational approach using human genetics and pharmacology

Open access
Liu, Xin, Chen, Yen-Wei, Xu, Xiao, Smith, David, Fan, Fan · 2026 · Toxicological Sciences

Abstract

Drug candidates are often evaluated for their activities against unexpected targets (off-targets), to either prospectively flag potential hazards or to provide mechanistic insights for a given phenotype. The in vitro to in vivo translatability is critical when selecting which "phenotypically consequential" off-targets to screen. To this end, human genetics and indication-based pharmacology offer unraveled insights. Enhanced natural language processing tools were applied to harness the power of large data obtained from 7 genetics and 2 pharmacology databases. Mapping biological roles to organ systems, we curated targets implicated in 22 organ systems of safety concerns, resulting in a safetyome composed of over ∼11,000 proteins. This is a significant expansion from our previously proposed screen, whose scope included phenotypes affecting 5 organ systems. Prioritization of the large panel using expression pattern and gene conservation across species resulted in a core panel of 500 targets. Mapping biological roles obtained from the databases to specific terms allowed us to systematically generate over 3,000 phenotype-based (specialized) panels, which can be used as gene or protein sets for issue resolution. All three components: The full safetyome, the core panel of 500 targets, and the over 3,000+ specialized panels, were systematically and orthogonally tested using independent data source, i.e., gene expression data from the Comparative Toxicogenomics Database. All panels, together with a user-friendly App, are published to aid effective safety assessment and issue resolution with strong "translational" focus.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1093/toxsci/kfag021

Cites

58 of 58 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1. Workflow for data harmonization and phenotype mapping. Data harmonization encompasses the curation of genetic and pharmacological data from nine distinct sources. Filtering criteria were established to refine associations based on statistical significance and source relevance. Phenotype mapping to the hierarchy employs NLP tools to systematically map terms or text to the MedDRA classification.
Fig. 1. Workflow for data harmonization and phenotype mapping. Data harmonization encompasses the curation of genetic and pharmacological data from nine distinct sources. Filtering criteria were established to refine associations based on statistical significance and source relevance. Phenotype mapping to the hierarchy employs NLP tools to systematically map terms or text to the MedDRA classification.CC BY · PubMed Central
Fig. 2. Distribution of unique targets across SOCs within safetyome. A) A bar chart depicting the number of unique genes associated with each SOC. (B) A circular chord diagram illustrating the interconnectivity between genes and their corresponding SOCs. C) Distribution of unique SOC counts across individual genes. The left y-axis indicates frequency, while the right y-axis shows percentage. D) A Venn diagram showing the overlap between genes derived from genetics and those from pharmacological datasets. E) An UpSet plot depicting the distribution of gene-SOC associations by data source. The bar chart in the top left quantifies the number of associations from individual sources, while the intersection matrix below highlights the overlap and unique gene-SOC pairs across databases. The pie chart in the upper right corner indicates the percentage of associations originating from multiple versus single sources.
Fig. 2. Distribution of unique targets across SOCs within safetyome. A) A bar chart depicting the number of unique genes associated with each SOC. (B) A circular chord diagram illustrating the interconnectivity between genes and their corresponding SOCs. C) Distribution of unique SOC counts across individual genes. The left y-axis indicates frequency, while the right y-axis shows percentage. D) A Venn diagram showing the overlap between genes derived from genetics and those from pharmacological datasets. E) An UpSet plot depicting the distribution of gene-SOC associations by data source. The bar chart in the top left quantifies the number of associations from individual sources, while the intersection matrix below highlights the overlap and unique gene-SOC pairs across databases. The pie chart in the upper right corner indicates the percentage of associations originating from multiple versus single sources.CC BY · PubMed Central
Fig. 3. Distribution of unique phenotype across SOCs and validation of phenotype-based gene sets. A) Number of unique phenotype-based gene sets associated with each SOC category. B) Match rates of phenotypes (PT level) enriched through statistical overrepresentation analysis of gene sets in CTD compared to the mapped phenotypes via NLP.
Fig. 3. Distribution of unique phenotype across SOCs and validation of phenotype-based gene sets. A) Number of unique phenotype-based gene sets associated with each SOC category. B) Match rates of phenotypes (PT level) enriched through statistical overrepresentation analysis of gene sets in CTD compared to the mapped phenotypes via NLP.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 54 references

Research works printed but not linked: 3

  1. Bonferroni, 1936, Teoria statistica delle classi e calcolo delle probabilita, Pubblicazioni Del R Istituto Superiore Di Scienze Economiche e Commericiali Di Firenze, 8, 3
  2. Kang, 2021, Paper presented at: Proceedings of the 2021 CHI conference on human factors in computing systems
  3. 2019, Scispacy: Fast and robust models for biomedical natural language processing

Cited by

See all 1

Metadata

Authors
Liu, X., Chen, Y.-W., Xu, X., Smith, D., & Fan, F.
Affiliations
Johnson & Johnson (United States) (US)
Corresponding author
Fan, Fan
Venue
Toxicological Sciences
Published
2026-02-19
Volume
209
Issue
3
ISSN
1096-0929 (ISSN-L)
OpenAlex
W7130606267
Cited by
1
References
54
Citations by year
2026: 1
FWCI
3.61
Type
journal article
Language
en
Open access
Locations
  • Toxicological Sciences (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
Funders
Johnson & Johnson
Topics
  • Life Sciences › Pharmacology, Toxicology and Pharmaceutics › Drug Discovery › Computational Drug Discovery Methods
  • Health Sciences › Medicine › Health Informatics › Biomedical Text Mining and Ontologies
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Bioinformatics and Genomic Networks
Keywords
off-target screening, drug safety assessment, human genetics, natural language processing, Comparative Toxicogenomics Database, gene expression, drug target prioritization, problem resolution
MeSH terms
Biocuration, Animals, Humans, Phenotype, Databases, Factual, Databases, Genetic, Toxicogenetics, Translational Research, Biomedical, Drug-Related Side Effects and Adverse Reactions

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Safetyome and specialized panels for over 3,000 phenotypes: a systematic and translational approach using human genetics and pharmacology. https://app.cochat.ai/papers/10.1093/toxsci/kfag021
The paper
Liu, X., Chen, Y.-W., Xu, X., Smith, D., & Fan, F. (2026). Safetyome and specialized panels for over 3,000 phenotypes: a systematic and translational approach using human genetics and pharmacology. Toxicological Sciences. https://doi.org/10.1093/toxsci/kfag021

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.