KNIndex

KNIndex provides a comprehensive database of physicochemical properties for k-tuple nucleotides to support genomic analyses, including genome annotation and machine learning-based analysis of DNA and RNA from high-throughput sequencing.

Key Features:

• Extensive data collection: Contains 182 distinct physicochemical properties covering mononucleotide (DNA), dinucleotide (147 DNA and 22 RNA), and trinucleotide (DNA) entries.
• Data aggregation: Consolidates physicochemical property values from multiple scattered resources into a unified dataset.
• Sequence-to-vector conversion: Provides mappings of k-tuple nucleotide properties to fixed-length numerical vectors for use in machine learning methods.
• Visualization functions: Includes functions to convert DNA and RNA sequences into visual representations such as curves of multiple physicochemical properties.

Scientific Applications:

• Genome annotation: Supplies physicochemical property features that can be used to enhance genome annotation workflows.
• Machine learning-based sequence analysis: Enables representation of sequences as fixed-length vectors to train and apply machine learning models for prediction tasks.
• Physicochemical characterization of sequences: Supports studies requiring detailed analysis of DNA and RNA physicochemical properties from high-throughput sequencing data.

Methodology:

Aggregation of physicochemical property values from multiple resources into a unified dataset; built-in functions to convert DNA/RNA sequences into curves of physicochemical properties; mapping of k-tuple nucleotide properties to fixed-length numerical vectors for machine learning.