Fig. 2

Overview of spectral and compound database dereplication performed in MAW-R. Spectral database dereplication utilizes HMDB, MassBank, and GNPS for each precursor mass [m/z]. The spectral databases are stored as Rdata objects, available on Zenodo. The spectra from input queries and databases with precursor mass [m/z] of interest are extracted. These spectra are compared with cosine similarity scores between the query and database spectra. Individual peaks are also compared to remove any false positive candidates with high overall similarity but lower individual peak similarity. The second part of MAW-R is the compound database dereplication with SIRIUS CLI. To prepare the input files for SIRIUS, MS² peaks and isotopic peaks are extracted. This information is written into SIRIUS input files (.ms), which are used to run SIRIUS