DUAL PARALLEL DETECTION RAW DATA FUSION: CHALLENGES AND OPPORTUNITIES FOR ACCURATE FINGERPRINTING OVER LARGE TIME FRAMES

Volatilomics is a growing field focused on identifying volatile metabolites in various samples. A straightforward approach using comprehensive two-dimensional gas chromatography (GC×GC) with parallel flame ionization detector (FID) and mass spectrometry (MS) offers both compound identification (via MS) and precise quantification (via FID). Typically, FID and MS data are analyzed separately, but with dedicated software solutions it is possible to combine raw signals from parallel detection. The fused signal brings the original information from both detectors (i.e., MS fragmentation pattern and FID response at each data point) and enable reliable image pattern recognition by template matching. The contribution discusses the workflow to obtain combined detector signals and challenges posed by: (a) dual parallel detection, (b) dual parallel second dimension (2D) columns set up, and (c) impaired acquisition frequencies. As test bench for the new data fusion workflow food volatilomics and fragrance allergens profiling are considered. With data fusion, template matching using MS spectral similarity reduces false negatives by 80% on comprehensive UT features covering the hazelnut detectable volatilome (450 UT features) and improves true positive matches in allergens recognition. In both applications, accurate quantification is made straightforward due to the readily available FID signal.