Qpalma

QPALMA: Quality-Aware Spliced Read Alignment Algorithm

QPALMA performs quality-aware spliced alignment of short sequencing reads to reference genomes by integrating base quality scores and computational splice site predictions to improve alignment accuracy across intron-exon junctions.

Key Features:

• Quality-Informed Alignment: Incorporates per-base sequencing quality scores into the alignment model to improve fidelity of spliced alignments.
• Splice Site Prediction Integration: Integrates computational splice site predictions to accurately map reads spanning intron-exon boundaries.
• Large Margin Parameter Estimation: Applies a support vector machine-inspired large margin approach trained on spliced reads with known alignments and quality information.
• Enhanced Suffix Array Mapping Pipeline: Combines alignment algorithms with a fast enhanced suffix array-based mapping pipeline for efficient processing of large next-generation sequencing datasets.

Scientific Applications:

• Transcriptome and Gene Structure Analysis: Enables identification of alternative splicing events, gene expression patterns, and novel transcripts through accurate alignment across intron boundaries in high-throughput sequencing data.

Methodology:

Optimized using reads generated by the Illumina Genome Analyzer and applied to Arabidopsis thaliana, with training on spliced reads containing known alignments and quality information to calibrate large margin model parameters.