Poster Open Access
Mass spectrometry proteomics: Ready for the deep learning (r)evolution?
Bittremieux, Wout; Laukens, Kris
In the past few years deep learning (DL) has revolutionized machine learning research, achieving tremendous increases in performance on a variety of problems, ranging from image recognition to natural language processing. In bioinformatics deep neural networks have already been used to solve important problems in genomics, but they have not seen a lot of use in biological mass spectrometry (MS) yet. Nevertheless, there is a huge opportunity to apply DL to MS research as well. Here we show how powerful DL models can usher in a shift from a model-driven approach, using hard-coded rules based on expert knowledge, for example such as complex fragmentation rules, to a data-driven one, where underlying rules are automatically inferred by advanced machine learning models.
| Name | Size | |
|---|---|---|
|
ASMS_2017_Mass_spectrometry_proteomics_Ready_for_the_deep_learning_(r)evolution.pdf
md5:e12fabb8f7b9f1e3758ed058c9708d67 |
2.4 MB | Download |
| All versions | This version | |
|---|---|---|
| Views | 136 | 136 |
| Downloads | 56 | 58 |
| Data volume | 135.9 MB | 140.7 MB |
| Unique views | 135 | 135 |
| Unique downloads | 48 | 50 |
- Publication date:
- June 8, 2017
- DOI:
-
- Meeting:
- ASMS conference, Indianapolis, IN, USA, 4-8 June 2017
- License (for files):
- Creative Commons Attribution Share Alike 4.0 International