StackPVP: Phage Virion Protein Predictions Using Evolutionary Features

StackPVP is a computational framework for predicting phage virion proteins (PVPs) using a stacked ensemble model. The dataset likely contains integrated evolutionary features derived from protein sequences, including PSSM, AAC-PSSM, DPC-PSSM, Pse-PSSM, and PSSM-COM. It was authored by Lixi Cai and last updated on March 18, 2026.

Use Cases

• Training classifiers for phage virion protein identification based on evolutionary features.
• Benchmarking new PVP prediction methods against the reported 94.26% AUC model.
• Feature engineering studies using integrated descriptors like PSSM-COM and Pse-PSSM.

Strengths

• The underlying model achieved an area under the curve (AUC) of 94.26% on the test dataset.
• The description details a specific methodological pipeline involving three feature selection methods and 12 base classifiers.
• The dataset is shared under a permissive CC-BY-4.0 license.

Limitations

• The dataset is very small at 43.2 KB, indicating limited scope.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

Lixi Cai via figshare.

Collection Method

Features were computationally derived from protein sequences using position-specific scoring matrices (PSSM).

Freshness

Last updated 2026-03-18 06:57:48; freshness should be verified.