Inpactor2

Inpactor2 identifies and classifies long terminal repeat (LTR) retrotransposons in plant genomes to support accurate annotation and study of their diversity and evolutionary roles.

Key Features:

• Hybrid detection approach: combines deep learning algorithms with structure-based methodologies to detect LTR-retrotransposons.
• Partial sequence filtering: filters out partial LTR-retrotransposon sequences.
• Superfamily and lineage classification: classifies intact elements into superfamilies and lineages.
• Reference library construction: builds comprehensive LTR-retrotransposon reference libraries.
• Parallel computing: leverages multi-core processing and GPU architectures to accelerate execution.
• Performance metrics: reports high accuracy and F1-Score, sensitivity exceeding other tools by 28%, specificity second only to EDTA, demonstrated a 5-minute run time on the rice genome, and up to sevenfold speed improvement over alternatives.

Scientific Applications:

• Genome annotation: identification and cataloging of LTR-retrotransposons in plant genomes for annotation workflows.
• Reference dataset generation: rapid creation of LTR-retrotransposon reference libraries for downstream analyses.
• Comparative and evolutionary genomics: genome-scale analyses of LTR-retrotransposon dynamics, evolution, and biodiversity in plant species.

Methodology:

Inpactor2 combines deep learning algorithms and structure-based methodologies to detect LTR-retrotransposons, filter partial sequences, classify intact elements into superfamilies and lineages, and construct reference libraries using multi-core and GPU processing.