An automated workflow for MALDI-ToF mass spectra pattern identification on large data sets: An application to detect aneuploidies from pregnancy urine

Abstract

Urine from first trimester pregnancies has been found to be rich in information related to aneuploidies and other clinical conditions. Mass spectral analysis derived from matrix assisted laser desorption ionization (MALDI) time of flight (ToF) data has been proven to be a cost effective method for clinical diagnostics. However, urine mass spectra are complex and require data modelling frameworks. Therefore, computational approaches that systematically analyse big data generated from MALDI-ToF mass spectra are essential. To address this issue, we developed an automated workflow that successfully processed large data sets from MALDI-ToF which is 100-fold faster than using a common software tool. Our method performs accurate data quality control decisions, and generates a comparative analysis to extract peak intensity patterns from a data set. We successfully applied our framework to the identification of peak intensity patterns for Trisomy 21 and Trisomy 18 gestations on data sets from maternal pregnancy urines obtained in the UK and China. The results from our automated comparative analysis have shown characteristic patterns associated with aneuploidies in the first trimester pregnancy. Moreover, we have shown that the intensity patterns depended on the population origin, gestational age, and MALDI-ToF instrument.