coil

coil assesses and preprocesses cytochrome c oxidase I (COI) 5' region (COI-5P) DNA barcode sequences to detect insertion and deletion (indel) errors and prepare sequences for downstream analyses.

Key Features:

• Data cleaning and contextualization: Provides functions for cleaning and contextualizing COI-5P DNA barcode sequences prior to analysis.
• Reading frame alignment: Aligns barcode sequences into a common reading frame to support accurate translation and error detection.
• Translation to amino acids: Translates DNA sequences into amino acids to facilitate identification of biological variation and sequence errors.
• Error assessment with profile HMMs: Employs nucleotide and amino acid profile hidden Markov models (HMMs) to assess the likelihood of indel errors in barcode sequences.
• Taxonomically diverse HMM training: Uses HMMs trained on a taxonomically diverse set of reference sequences to improve error-detection performance.
• Error identification accuracy: Demonstrates over 97.5% accuracy in distinguishing error-free sequences from those containing indel errors across an analysis of 10,000 barcode sequences.

Scientific Applications:

• Distinguishing biological variation from experimental error: Aids interpretation of DNA barcoding and metabarcoding data by separating technical indel errors from biological sequence variation.
• Limitations and specificity: Applicable primarily to animal COI-5P sequences and not designed to detect all contaminants, with reduced reliability for non-animal kingdoms, nuclear mitochondrial pseudogenes, and Wolbachia COI-5P sequences.

Methodology:

Aligns sequences into a common reading frame, translates DNA to amino acids, and compares input sequences against trained nucleotide and amino acid profile HMMs (trained on taxonomically diverse reference sequences) to evaluate the presence of technical indel errors, following approaches described by Durbin et al. (1998).