Co-occurrence network multicollinearity artefacts using centered log-ratio transformation
Next-generation sequencing data is compositional in nature and ignoring this fact can significantly impact downstream data analyses. This is especially true in co-occurrence (correlation) networks because compositional data is prone to many false negative correlations. The centered log-ration (clr) transformation is a commonly used compositional data analysis (CoDA) approach that converts…
Co-occurrence networks derived from next-generation sequencing data are susceptible to multicollinearity artefacts due to compositional data properties. Ignoring this fact can lead to misleading results in downstream analyses. The centered log-ratio (clr) transformation is a widely used method to address this issue by converting constrained compositional data into real number space.
This transformed data can then be utilized in statistical analyses, such as Spearman or Pearson correlations, to generate co-occurrence networks. However, the conventional clr transformation involves a zero substitution step, which has been found to introduce numerous multicollinearity artefacts in correlation-based network analysis.
In contrast, a newer CoDA approach called the robust centered log-ratio (rclr) has been shown to effectively eliminate these artefacts, making it a preferable alternative for generating accurate co-occurrence networks.
Written by urgent.news from bioRxiv's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
