SpindleP

SpindleP predicts whether human genes encode proteins associated with the meiotic and mitotic spindle apparatus to identify spindle-associated components and hubs involved in cell division.

Key Features:

• Integrated Computational Approach: Integrates data from multiple sources to predict potential spindle proteins by analyzing predicted relationships among candidates.
• Neural Network Prediction: Uses neural networks to estimate the likelihood that human genes encode spindle-associated proteins.
• Experimental Validation: Facilitates experimental confirmation of predicted proteins' association with the spindle and roles in mitosis.
• High Success Rate: Reported experimental validation shows a 75% success rate for SpindleP predictions versus 35% using expert knowledge alone.
• Identification of Hidden Spindle Hubs: Predicts proteins with poorly characterized interaction networks that may act as hidden hubs influencing spindle functionality.

Scientific Applications:

• Discovery of novel spindle components: Identifies previously uncharacterized proteins associated with the mitotic and meiotic spindle for further study.
• Characterization of hidden hubs: Highlights proteins with sparse interaction data that may have large-scale effects on spindle organization.
• Prioritization for experimental validation: Ranks candidate spindle proteins to guide follow-up cellular biology, high-throughput proteomics, and RNA interference experiments.

Methodology:

Integrates high-throughput proteomics and RNA interference experimental data with computational predictions using neural networks and analysis of predicted relationships among proteins.