2020
DOI: 10.1016/j.csbj.2020.09.033
|View full text |Cite
|
Sign up to set email alerts
|

Deep metabolome: Applications of deep learning in metabolomics

Abstract: Graphical abstract

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
67
0

Year Published

2021
2021
2024
2024

Publication Types

Select...
8
1
1

Relationship

0
10

Authors

Journals

citations
Cited by 110 publications
(67 citation statements)
references
References 84 publications
(100 reference statements)
0
67
0
Order By: Relevance
“…Thus, as second approach, we used random forest (no. of trees = 5000), a machine learning approach 20 to select the best performing lipid species per pairwise comparison, based on the lowest mean values for minimum depths in the trees (lower the better) and the frequencies found in trees (higher the better). Minimum depth indicates how early in decision trees a lipid species is involved (Fig.…”
Section: Resultsmentioning
confidence: 99%
“…Thus, as second approach, we used random forest (no. of trees = 5000), a machine learning approach 20 to select the best performing lipid species per pairwise comparison, based on the lowest mean values for minimum depths in the trees (lower the better) and the frequencies found in trees (higher the better). Minimum depth indicates how early in decision trees a lipid species is involved (Fig.…”
Section: Resultsmentioning
confidence: 99%
“…Despite the considerable efforts of the aforementioned ML methods to rapidly and accurately characterize non-linear complex samples and their metabolite properties, false positive signals, co-eluting metabolites and retention time shifts are still major bottlenecks that effect data analysis and interpretation of plant MS-based metabolomics. Artificial neural networks (ANNs) and deep learning (DL) methods are proposed to solve these issues and other bottlenecks involved in the mining of metabolomics data [ 154 , 197 ].…”
Section: 4ir Technologies and Plant Metabolomicsmentioning
confidence: 99%
“…Artificial neural networks are ML tools based on interconnecting hidden layers, computational structures inspired by neurons in the brain, and in their simplest forms are similar to PLS but can model nonlinear models. Deep neuronal networks, or DL, can predict relationships from diverse datasets and can accomplish supervised, semisupervised and unsupervised tasks, improving the interpretability of data analysis [ 140 , 141 , 142 , 143 ]. DL techniques transform the data by iteratively tuning their internal parameters and may enable the extraction of the most predictive features from complex datasets.…”
Section: Machine Intelligence and Learning Approachesmentioning
confidence: 99%