iBLP

iBLP predicts bioluminescent proteins (BLPs) from protein sequences using eXtreme Gradient Boosting (XGBoost) and sequence-derived features to identify BLPs across bacteria, eukaryotes, and archaea.

Key Features:

• XGBoost classifier: Employs the eXtreme Gradient Boosting (XGBoost) algorithm for classification of bioluminescent proteins.
• Sequence-derived features: Utilizes features extracted from protein sequences as input for prediction.
• Taxonomic coverage: Models and datasets include BLPs sourced from bacteria, eukaryotes, and archaea.
• Feature and algorithm evaluation: Systematically evaluates various feature extraction methods and classification algorithms to optimize performance.
• Validation on multiple datasets: Performs testing on training and independent datasets to assess robustness.
• Comparative analysis: Includes comparisons with existing methodologies to assess relative predictive capability.

Scientific Applications:

• Gene expression analysis: Supports identification of BLPs relevant for use as reporters in gene expression studies.
• Drug discovery: Aids in identifying BLPs that can be applied in screening or assay development within drug discovery workflows.
• Cellular imaging: Facilitates detection of BLPs for applications in cellular and molecular imaging techniques.
• Toxicity determination: Assists in identifying BLPs applicable to assays for toxicity assessment.

Methodology:

Uses eXtreme Gradient Boosting (XGBoost) on sequence-derived features, with systematic evaluation of feature extraction methods and classification algorithms, and validation on training and independent datasets including comparative analyses with existing methods.