Deep learning of 2D-Restructured gene expression representations for improved low-sample therapeutic response prediction.

Clinical outcome prediction is important for stratified therapeutics. Machine learning (ML) and deep learning (DL) methods facilitate therapeutic response prediction from transcriptomic profiles of cells and clinical samples. Clinical transcriptomic DL is challenged by the low-sample sizes (34-286 subjects), high-dimensionality (up to 21,653 genes) and unordered nature of clinical transcriptomic data. The established methods rely on ML algorithms at accuracy levels of 0.6-0.8 AUC/ACC values. Low-sample DL algorithms are needed for enhanced prediction capability. Here, an unsupervised manifold-guided algorithm was employed for restructuring transcriptomic data into ordered image-like 2D-representations, followed by efficient DL of these 2D-representations with deep ConvNets. Our DL models significantly outperformed the state-of-the-art (SOTA) ML models on 82% of 17 low-sample benchmark datasets (53% with >0.05 AUC/ACC improvement). They are more robust than the SOTA models in cross-cohort prediction tasks, and in identifying robust biomarkers and response-dependent variational patterns consistent with experimental indications.

Publisher Correction: Non-Toxic Gold Nanoclusters for Solution-Processed White Light-Emitting Diodes.

A correction to this article has been published and is linked from the HTML and PDF versions of this paper. The error has been fixed in the paper.

Non-Toxic Gold Nanoclusters for Solution-Processed White Light-Emitting Diodes.

Solution-processed optoelectronic devices are attractive because of the potential low-cost fabrication and the compatibility with flexible substrate. However, the utilization of toxic elements such as lead and cadmium in current optoelectronic devices on the basis of colloidal quantum dots raises environmental concerns. Here we demonstrate that white-light-emitting diodes can be achieved by utilizing non-toxic and environment-friendly gold nanoclusters. Yellow-light-emitting gold nanoclusters were synthesized and capped with trioctylphosphine. These gold nanoclusters were then blended with the blue-light-emitting organic host materials to form the emissive layer. A current efficiency of 0.13 cd/A was achieved. The Commission Internationale de l'Eclairage chromaticity coordinates of (0.27, 0.33) were obtained from our experimental analysis, which is quite close to the ideal pure white emission coordinates (0.33, 0.33). Potential applications include innovative lighting devices and monitor backlight.