DECIPHER, a search-based approach to chimera identification for 16S rRNA sequences

Area of Science:

• Microbiology
• Bioinformatics
• Computational Biology

Background:

• Chimeric sequences are artifacts in 16S rRNA gene sequencing that can distort microbiome analyses.
• Accurate identification of chimeras is crucial for reliable phylogenetic and ecological interpretations.

Purpose of the Study:

• To introduce DECIPHER, a new search-based algorithm for detecting 16S rRNA chimeric sequences.
• To evaluate DECIPHER's performance against existing chimera detection tools.

Main Methods:

• DECIPHER was developed using a search-based approach to identify chimeric fragments.
• The algorithm was calibrated for both full-length (fs_DECIPHER) and short sequences (ss_DECIPHER).
• Performance was benchmarked against WigeoN, ChimeraSlayer, and Uchime using simulated chimeric sequences.

Main Results:

• ss_DECIPHER demonstrated high chimera detection rates (89%) for sequences 100-600 nucleotides, outperforming Uchime for longer sequences.
• fs_DECIPHER achieved a 75% detection rate for longer sequences with the lowest false-positive rate (<0.20%).
• Combining ss_DECIPHER and Uchime enabled detection of over 89% of all chimeras; fs_DECIPHER identified additional chimeras in public databases.

Conclusions:

• DECIPHER provides an effective and accurate method for identifying 16S rRNA chimeric sequences.
• The tool, available as an R package and webpage, enhances the quality of microbiome datasets.
• DECIPHER contributes to more reliable downstream analyses in microbial ecology and evolution.