Statistical t-test p-values for PepLM-GNN Model Comparison

PepLM-GNN's statistical superiority is evaluated through t-test p-values against other peptide prediction methods. The data aggregates results from four independent test sets, using mean ACC values from five-fold cross-validation. A p-value below 0.05 indicates PepLM-GNN's significant performance advantage.

Use Cases

• Compare PepLM-GNN's ACC performance to other methods by analyzing the reported p-values.
• Validate the statistical significance of the hybrid graph network's results across the four combined test sets.
• Assess the generalisation performance of peptide prediction models using the provided t-test metrics.

Strengths

• Results are aggregated from four independent test datasets, increasing robustness.
• Statistical tests are based on mean ACC values from a five-fold cross-validation procedure.
• The dataset clearly defines a p-value threshold of <0.05 for establishing statistical superiority.

Limitations

• The dataset is very small at 5.5 KB, containing only summary statistical results, not raw prediction data.
• No column names or sample data are provided, limiting direct analytical utility.
• The data only covers a specific comparison for PepLM-GNN, not a broad benchmark of methods.

Provenance

Source

figshare, authored by Ke Yan.

Collection Method

Derived from statistical t-tests on mean ACC values from five-fold cross-validation across four combined test sets.

Time Range

null

Freshness

null

Geography

null