Exploring the potential of MALDI imaging mass spectrometry for personalized biomarker analysis in triple-negative breast cancer patients.

点击翻译按钮获取中文摘要

The enormous size of data accumulated by employing immunohistochemistry, ELISA and Q-RT-PCR, plus MALDI Imaging Mass Spectrometry (MALDI IMS), acquired within the joint project on biomarker analysis in triple-negative breast cancer, makes automated analysis indispensable. We thus propose to develop and deploy a suite of different state-of-the-art methods from statistics, machine learning, and image processing where a special focus lies on the following six areas of application: (1) robust preprocessing of raw IMS data (baseline correction, normalization, peak picking, peak alignment), (2) comparison of current approaches for the classification of different tissue types across patients based on MALDI IMS signatures (e.g. automated distinction of triple-negative breast cancer tissue from triple-positive controls and connective tissue), (3) prediction of therapy response to cancer therapeutics as a precondition for personalized medicine, (4) prediction of patient disease-free and overall survival times, (5) confirmation of established as well as identification of novel biomarkers in these settings, and (6) development of a graphical annotation and classification tool for analysis of imaging mass spectrometry data.