Figure - available via license: Creative Commons Attribution-NonCommercial 4.0 International
Content may be subject to copyright.
Download
View publication
Figur e 3: T he first ML integr ation fr ame work. ( A ) Multi-omics fusion in conv entional ML. The first m ulti-omics model in conv entional ML outperforms any single-omics model in terms of AUC, with an increase of 0.06-0.17. ( B ) Permutation test. Permutation testing (100 permutations) in four ML methods shows a significant difference between the original AUC and random permutations ( P < 0.05). The mean AUC area is 0.50-0.52, and the variance is 0.01-0.02. ( C ) Single-omics model in conventional ML. DT, decision tree; RF, random forest.

Figur e 3: T he first ML integr ation fr ame work. ( A ) Multi-omics fusion in conv entional ML. The first m ulti-omics model in conv entional ML outperforms any single-omics model in terms of AUC, with an increase of 0.06-0.17. ( B ) Permutation test. Permutation testing (100 permutations) in four ML methods shows a significant difference between the original AUC and random permutations ( P < 0.05). The mean AUC area is 0.50-0.52, and the variance is 0.01-0.02. ( C ) Single-omics model in conventional ML. DT, decision tree; RF, random forest.

Source publication
Figure 1: Scheme of feature fusion evaluation. ( A ) The first ML...
Figure 4: Contributions of features. ( A ) Abnormal brain regions of...
Figur e 3: T he first ML integr ation fr ame work. ( A ) Multi-omics...
AUC of the first ML integration framework.
Neural network results.
Transcriptomic and Neuroimaging Data Integration Enhances Machine Learning Classification of Schizophrenia
Article
Full-text available
  • Mar 2024
Background Schizophrenia is a polygenetic disorder associated with changes in brain structure and function. Integrating macroscale brain features and microscale genetic data may provide a more complete overview of the disease etiology and may serve as potential diagnostic markers for schizophrenia. Objective We aim to systematically evaluate the i...
Cite
Download full-text

Context in source publication

Context 1
... first illustrate the classification performance (i.e. AUC) of the integr ated featur e fusion in contrast to the performance ac hie v ed by each single modality (Fig. 3 ). The number of features in the classifier are 17,19,74,150,45, and 156, separ atel y. Corr esponding outcomes across these varying feature sets are detailed in Supplementary Figs. 2 -4 and Supplementary Tables 4 -6 , showing similar performance to what we report in the main text. For all classifiers, the first multi-omics fusion model ...