Selenoprofiles3

Selenoprofiles3 predicts and annotates eukaryotic selenoprotein genes by detecting selenocysteine (Sec) encoded at UGA codons using homology-based searches against curated multiple sequence alignments.

Key Features:

• Homology-Based Prediction: Utilizes manually curated multiple sequence alignments of known selenoprotein families as queries to scan eukaryotic genomic sequences.
• UGA Codon Resolution: Specifically addresses the dual functionality of UGA codons to distinguish selenocysteine (Sec) insertion from translational termination.
• Integrated Analysis Pipeline: Integrates psitblastn, exonerate, genewise, SECISearch, and custom scripts to process scan results and generate gene predictions.
• SECIS Detection: Uses SECISearch to identify SECIS elements required for Sec insertion in eukaryotic transcripts.
• High Accuracy and Reliability: Leverages curated alignments and downstream processing steps to produce precise selenoprotein gene predictions.
• Minimal Human Intervention: Automates core prediction and processing steps to reduce manual input and increase throughput.
• Implementation: Implemented in Python.

Scientific Applications:

• Selenoprotein Discovery: Identification of Sec-containing genes across eukaryotic genomes.
• Genome Annotation: Complementation of existing genome annotation pipelines to enhance characterization of eukaryotic selenoproteomes.
• Error Correction: Resolution of misannotations caused by UGA codons being annotated solely as stop codons.

Methodology:

Employs a homology-based in silico approach using manually curated multiple sequence alignments to scan genomes with psitblastn, constructs gene models with exonerate and genewise, detects SECIS elements with SECISearch, and processes results with custom scripts.