Performance of a Shotgun Prediction Model for Colorectal Cancer When Using 16S rRNA Sequencing Data

Area of Science:

• Microbiome Research
• Cancer Genomics
• Bioinformatics

Background:

• Colorectal cancer (CRC) is a leading global cancer, with gut microbiota dysbiosis implicated in its development.
• Shotgun metagenomic sequencing and 16S ribosomal RNA (16S) gene sequencing are key methods for microbiome analysis, but yield different results.
• Validating CRC-associated microbial signatures between these sequencing methods presents a significant challenge due to data discrepancies.

Purpose of the Study:

• To develop and evaluate an algorithm for mapping shotgun metagenomic sequencing taxa to 16S ribosomal RNA (16S) gene sequencing counterparts.
• To enable the assessment of shotgun-derived microbiome signatures for CRC prediction using 16S data.
• To facilitate comparative analysis and validation of microbiome data across different sequencing techniques in CRC research.

Main Methods:

• An algorithm was created to map taxonomic profiles obtained from shotgun metagenomic data to corresponding 16S rRNA gene sequences.
• The predictive performance of a CRC microbiome signature, initially derived from shotgun data, was evaluated using the 16S-mapped taxa.
• Statistical significance was assessed to determine the reliability of the 16S-mapped signature in prediction models.

Main Results:

• The application of 16S-mapped taxa in the shotgun prediction model resulted in a decrease in predictive performance compared to the original shotgun signature.
• Despite the performance reduction, the 16S-mapped signature retained statistical significance for CRC prediction.
• The study highlights the feasibility of cross-platform validation, even without perfect data congruence between shotgun and 16S sequencing.

Conclusions:

• The developed algorithm provides a viable approach for comparing and validating microbiome signatures between shotgun and 16S sequencing data in colorectal cancer research.
• While direct equivalence is not achieved, the method allows for meaningful cross-validation, supporting the use of 16S data for validating shotgun-derived findings.
• This work contributes to advancing the understanding of the gut microbiome's role in CRC and standardizing analytical approaches.