ConoDictor2

ConoDictor2 classifies conopeptides from cone snail venom gland transcriptomes using Hidden Markov Models (HMMs) and generalized profiles to assign amino acid sequences to gene superfamilies.

Key Features:

• Superfamily classification: Classifies conopeptide amino acid sequences into gene superfamilies using Hidden Markov Models (HMMs) and generalized profiles.
• Enhanced profiles and models: Employs updated classification models and enhanced generalized profiles trained on high-quality sequences to improve prediction accuracy.
• Algorithmic performance: Incorporates algorithmic improvements that enable processing of whole venom gland transcriptomes with reported 4–8× speed improvements on a single-core computer.
• Input formats and scale: Processes raw reads and contigs from venom gland transcriptomes for large-scale analyses.
• Novel and distant peptide detection: Improves identification and classification of novel or distantly related conopeptides.
• Benchmarking and comparison: Validated against public conopeptide databases and specific venom duct transcriptomes and compared with ConoSorter and BLAST.

Scientific Applications:

• Mining novel conopeptides: Facilitates identification of novel conopeptides from venom gland transcriptomes.
• Pharmacology and therapeutics: Supports discovery and development of bioactive peptides for pharmacological and therapeutic research.

Methodology:

Uses Hidden Markov Models (HMMs) and generalized profiles with updated classification models trained on high-quality sequences; processes raw reads or contigs from venom gland transcriptomes; validated against public conopeptide databases and venom duct transcriptomes including Conus geographus and compared with ConoSorter and BLAST.