yamss

The 'yamss' software tool addresses mass spectrometry-based metabolomics data analysis challenges, particularly in biomedical research. The existing paradigm of preprocessing metabolomics data involves extracting features from each sample separately and then attempting to group these features across samples for comparison. However, this approach introduces unnecessary variability in peak quantifications, adversely affecting downstream analyses. 'yamss' introduces a novel method called bakedpi for preprocessing centroid and profile mode metabolomics data. The method relies on intensity-weighted bivariate kernel density estimation to pool all samples to detect peaks. Doing so reduces unnecessary quantification variability, leading to increased power in downstream differential analysis.