SPALN

SPALN maps and aligns cDNA or amino acid sequences onto eukaryotic genomic sequences with spliced alignment to identify gene structures.

Key Features:

• Protein query support: SPALN accepts protein sequences as queries in addition to cDNA sequences, enabling the use of distantly related homologs as templates for gene recognition.
• Self-sufficient genome mapping: SPALN performs genome mapping and spliced alignment without relying on external tools such as Blast search followed by dynamic programming.
• Species-specific scoring: Spaln2 incorporates additional species-specific features in its scoring scheme, improving accuracy for plants and fungi and for increasingly divergent pairs of target and query sequences.
• Speed optimization: Spaln3 introduces the multi-intermediate unidirectional Hirschberg method and SIMD-based vectorization, achieving approximately 7-fold speed increase over Spaln2.
• Gene-structure-aware MSA construction: SPALN cooperates with Prrn5 to construct gene-structure-aware multiple sequence alignments (GSA-MSAs) from genomic sequences and transcript sequences without requiring external tools or annotation files.

Scientific Applications:

• Gene structure prediction: SPALN identifies eukaryotic gene structures including exon-intron boundaries through spliced alignment of transcript or protein queries to genomic DNA.
• Cross-species gene annotation: SPALN enables gene prediction in newly sequenced genomes using protein sequences from distantly related species as templates, with higher accuracy than conventional Blast-based approaches.
• Benchmarking spliced alignment: SPALN has been evaluated on simulated sequence datasets and actual cross-species datasets designed to represent various real situations including divergent eukaryotic species and wide divergence between query and target sequences.
• Gene prediction error detection: GSA-MSAs constructed with SPALN and Prrn5 detect and remedy gene-prediction errors prevalent in genome annotations produced by various genome projects.
• Evolutionary analysis of gene structure: GSA-MSA analysis with SPALN examines gain and loss of introns during the evolution of homologous eukaryotic genes.

Methodology:

SPALN uses dynamic programming for spliced alignment. Spaln3 speeds up the dynamic programming calculation through the multi-intermediate unidirectional Hirschberg method and SIMD-based vectorization.