ClanTox

ClanTox classifies toxin-like proteins across animal species by applying a machine learning classifier trained on known animal toxins to analyze primary amino acid sequences.

Key Features:

• Toxin Classification System: Extracts sequence-driven features from primary amino acid sequences and determines the likelihood that a sequence is toxin-like.
• Broad Applicability: Processes protein sequences from venomous and non-venomous organisms, including scorpions, snakes, bees, wasps, frogs, stonefish, snails, jellyfish, and hydra.
• Detection of Undetected Proteins: Identifies toxin-like proteins that are often overlooked by traditional search engines and genome annotation tools due to short length and primary sequence diversity.
• Statistical Confidence Ranking: Provides a ranked list of positively predicted candidates based on statistical confidence levels.
• Additional Protein Information: Reports features such as presence of signal peptides, number of cysteine residues, and associated functional annotations.

Scientific Applications:

• Discovery and Prediction: Detects previously overlooked toxin-like proteins, including short and sequence-diverse peptides, to expand toxin inventories.
• Therapeutic and Mechanistic Research: Aids identification of potential therapeutic agents among toxin-like cell modulators and supports studies of venomous mechanisms and novel medical treatments.

Methodology:

Extracts features from protein sequences and applies a machine learning-based classification system trained on an extensive dataset of known animal toxins.