Prospectr

Prospectr prioritizes candidate genes within large genomic linkage regions (>30 centimorgans) by analyzing sequence-based features to rank genes for their likelihood of involvement in human hereditary disease.

Key Features:

• Sequence-based feature analysis: Leverages sequence-based features that distinguish disease-related genes from non-disease genes in human hereditary diseases.
• Machine learning classifier: Employs an alternating decision tree algorithm trained on known disease and non-disease gene sets to automatically prioritize candidate genes.
• Enrichment capability: Enriches candidate lists, increasing the proportion of disease-related genes by two-fold 77% of the time, five-fold 37% of the time, and twenty-fold 11% of the time.
• Performance: Outperforms existing sequence-based classifiers when applied to novel data.
• Large-region prioritization: Designed to prioritize positional candidates within expansive linkage regions that can contain hundreds of genes.

Scientific Applications:

• Mendelian and oligogenic disorder gene discovery: Prioritizes genes involved in Mendelian (single-gene) and oligogenic (few-gene) disorders to guide mutation detection and case-control association studies.
• Positional candidate reduction in linkage studies: Focuses follow-up experimental efforts by reducing candidate gene lists from large genomic regions identified by genetic linkage.

Methodology:

Analyzes sequence-based features and applies an alternating decision tree machine learning algorithm to rank genes by their likelihood of disease involvement.