Non-convex quasi-norm-based normalization of MALDI-MS Imaging Data
Normalization is used as a pre-step before applying multivariate techniques or to improve visibility in MALDI imaging datasets. Total ion count (TIC) normalization calculates the normalization factor for each pixel as the l1-norm of the signal intensities. It has been found to be less prone to artefacts compared to variations based on higher norms (i.e., l2 or RMS, l∞ or maxima-based). We investigate the use of generalized lp-quasi-norms, with 0 < p < 1. As p gets smaller, the increasing influence of smaller signal intensities in the normalization factor tend to be even less prone to artefacts, but normalization ranges become less uniform across pixels. A reasonable trade-off is proposed within 0.75 ? p ? 1.
- Content Type:
- Paper
- Document Number:
- PO-CON1559E
- Product Type:
- Mass Spectrometry, Life Science Lab Instruments, MALDI-TOF Mass Spectrometry
- Keywords:
- Non-convex quasi-norm-based, angiotensin peptide, Pharmaceutical, Life Science, MALDI-MS Imaging, MALDI-7090
- Language:
- English
- File Name:
- uko315071.pdf
- File Size:
- 693kb