Rolexa

Rolexa implements a model-based base-calling algorithm for Solexa/Illumina sequencing that identifies and encodes ambiguous bases to select optimal sub-tags and increase usable short-read tags and genome coverage.

Key Features:

• Base-calling algorithm: Uses model-based clustering and probability theory to perform base calling from Solexa/Illumina data.
• Ambiguity coding: Identifies ambiguous bases within reads and encodes them using IUPAC ambiguity symbols.
• Information-content scoring: Applies a scoring system based on information content to select optimal sub-tags and remove uncertain bases.
• End trimming: Preferentially removes uncertain bases located at the ends of sequences to improve tag usability.
• Read recovery: Increases the number of usable tags and genome coverage by an average of ~15% compared to the standard Solexa pipeline.
• Input data: Processes Solexa/Illumina fluorescence intensity files from sequencing-by-synthesis platforms.
• Implementation: Provided as an R package for fast base calling from fluorescence intensity files.
• Diagnostic outputs: Generates diagnostic plots for visualization and interpretation of sequencing data quality and processing outcomes.

Scientific Applications:

• Short-read base calling: Base calling for high-throughput Solexa/Illumina short-read datasets (reads up to 36 bases).
• Tag recovery: Recovery of ambiguous or low-confidence tags to maximize usable tag yield and effective throughput.
• Genome coverage improvement: Increasing usable tags and genome coverage to improve the accuracy of downstream genomic analyses.
• Quality assessment: Visualization and interpretation of sequencing data quality via diagnostic plots derived from fluorescence intensity files.

Methodology:

Rolexa applies model-based clustering and probability-theory–based base calling, encodes ambiguous bases with IUPAC symbols, scores sub-tags by information content to select optimal sub-tags and trim uncertain terminal bases, and operates on Solexa/Illumina fluorescence intensity files as implemented in an R package.