Multiscale peak detection in wavelet space

摘要: Accurate peak detection is essential for analyzing high-throughput datasets generated by analytical instruments. Derivatives with noise reduction and matched filtration are frequently used, but they sensitive to baseline variations, random deviations in the shape. A continuous wavelet transform (CWT)-based method more practical popular this situation, which can increase accuracy reliability identifying peaks across scales space implicitly removing as well baseline. However, its computational load relatively high estimated features of may not be accurate case that overlapping, dense or weak. In study, we present multi-scale (MSPD) taking full advantage additional information including ridges, valleys, zero-crossings. It achieve a thresholding each detected maximum ridge. has been comprehensively evaluated MALDI-TOF spectra proteomics, CAMDA 2006 SELDI dataset Romanian database Raman spectra, particularly suitable detecting signals. Receiver operating characteristic (ROC) curves show MSPD detect true while keeping false discovery rate lower than MassSpecWavelet MALDIquant methods. Superior results suggest seems universal detection. designed implemented efficiently Python Cython. available an open source package at .