WGS Data for ML Prediction of Antibiotic Resistance in E. coli

1952 Escherichia coli isolates with whole-genome sequencing data and antibiotic susceptibility test phenotypes. The dataset was used to train machine learning classifiers for predicting resistance to 10 common antibiotics, with model performance evaluated using AUC and other metrics. Author Fang Wan published the data on figshare under a CC-BY-4.0 license, last updated on 2026-04-28.

Use Cases

• Training ML models to predict antibiotic resistance based on gene features extracted from WGS data.
• Comparing the performance of gene-based, SNP-based, and k-mer feature representations for AMR prediction tasks.
• Identifying core resistance genes, such as aac(3)-IIg or blaTEM, through feature-importance analysis.
• Developing clinical decision-support tools for rapid phenotype prediction from genomic sequences.

Strengths

• Includes data from 1952 E. coli isolates, providing a substantial sample size for model training.
• Model performance is quantified with specific AUC scores ranging from 0.6691 to 0.9879 across 10 antibiotics.
• Feature-importance analysis identified 40 core genes concordant with established resistance mechanisms.

Limitations

• The dataset is small in file size (487.0 KB), suggesting limited raw data or a processed summary.
• Column-level documentation and sample data are unavailable, requiring manual inspection after download.
• Row count is unknown, which may limit suitability assessment for specific modeling tasks.

Provenance

Source

figshare

Collection Method

Analysis of whole-genome sequencing data from bacterial isolates with associated antibiotic susceptibility testing.

Freshness

Last updated 2026-04-28 05:24:13; freshness should be verified.