Resumen de Statistical evaluation of multi-template PCR biases in microbiome data
Ilia Korvigo, Evgeny E. Andronov
High-throughput sequencing (HTS) of marker gene libraries has enabled microbiome research on an unprecedented scale and has become the de facto standard in the field. Although the method is renowned for high qualitative phylogenetic sensitivity and accessibility, owing to relative methodological simplicity and low sequencing costs, its quantitative prowess remain an open question in the face statistical properties of the data it generates, namely their compositional nature. Several normalisation and calibration methods have been proposed to both augment marker gene sequencing data with scale and lift compositional constraints, thereby enabling conventional quantitative statistical analysis, though none of them have addressed critical practical issues, such as multi-template PCR bias. Seeing how PCR biases might hamper qualitative microbiome research, we have designed and carried out a study to elucidate temporal dynamics of community compositions undergoing multi-template PCR and, hopefully, provide a prospect of analytical solution in case the problem proves to be too severe to neglect. Our results prove that we cannot carry out reliable quantitative microbiome research without accounting for PCR biases.
