LOCUS

LOCUS predicts gene structures and characterizes alternative spliceforms by integrating length constraints into a dynamic programming framework to improve splice-site and splice-variant identification.

Key Features:

• Dynamic Programming Framework: Employs a dynamic programming approach to identify optimal sets of splice sites and gene structures within genomic regions.
• Length Constraint Integration: Incorporates length constraints derived from prior information about expected transcript sizes to improve prediction of complete transcripts and alternative splicing patterns.
• Evidence-Based Prediction: Leverages multiple evidence sources including sequence conservation, known protein matches, and partial cDNA sequences to corroborate gene and spliceform predictions.
• High-Throughput Compatibility: Accepts length information that can be obtained in a high-throughput manner for integration into large-scale gene-prediction workflows.

Scientific Applications:

• Alternative splicing prediction: Predicts alternative splice variants and splice sites, with emphasis on accurately characterizing spliceforms.
• Complete gene structure reconstruction: Predicts complete gene structures and mRNA products to support comprehensive mRNA set characterization.
• Benchmarking and validation: Demonstrated superior performance versus ab initio gene finders and was validated on a test set of alternatively spliced internal exons from Caenorhabditis elegans.
• Functional genomics and transcriptomics: Supports analyses aimed at elucidating functional genomic and transcriptomic complexity by improving splice-variant annotation.

Methodology:

Integrates length information into a dynamic programming framework to predict gene products and spliceforms; uses evidence-based prediction by incorporating sequence conservation, known protein matches, and partial cDNA sequences; accepts length information obtainable in high-throughput formats.