BATH

BATH annotates protein-coding DNA based on direct alignment to protein sequences or profile hidden Markov models, incorporating frameshift-aware algorithms to detect nucleotide insertions and deletions.

Key Features:

• Frameshift-aware alignment: BATH introduces novel frameshift-aware algorithms to detect frameshift-inducing nucleotide insertions and deletions.
• HMMER3-based architecture: BATH is built on top of the HMMER3 code base for pHMM-based translated sequence annotation.
• Error-tolerant annotation: BATH matches HMMER3 accuracy for sequences without errors and produces superior accuracy for sequences containing nucleotide indels.

Scientific Applications:

• Long-read sequencing annotation: BATH should be used when high annotation sensitivity is required, particularly when frameshift errors are expected to interrupt protein-coding regions, as occurs with long-read sequencing data.
• Pseudogene analysis: BATH is applicable for annotation of protein-coding DNA in the context of pseudogenes where frameshift errors are expected.

Methodology:

BATH performs direct alignment of protein-coding DNA to a database of protein sequences or profile hidden Markov models, using frameshift-aware algorithms to detect frameshift-inducing nucleotide insertions and deletions.