In this paper, a nanobiosensor with surface-enhanced Raman scattering (SERS) capability is introduced for highly sensitive miRNA detection in colorectal cancer. This sensor was designed and fabricated by employing a nanoshielding mechanism from nanopolystyrene beads to resist reactive ion etching and allow anisotropic electrochemical etching, producing high-aspect-ratio, surface-corrugated nanopillars (SiNPs) on a silicon wafer to create extensive hot spots along the nanopillars for improved SERS signals. SERS enhancements were correlated with nanorange roughness, indicating that hot spots along the pillars were the crucial factor to improve the SERS effect. We achieved the detection capability of a trace amount of R6G (10⁻⁸ M), and the SERS signal enhancement factor (EF) was close to 1.0 × 10⁷ on surface-corrugated gold SiNPs. miRNA samples were also demonstrated on this sensor with good sensitivity and specificity. The target molecule miR-21-Cy5 was easily monitored through Raman spectrum variation with a PCR-comparable concentration at around 100 pM with clear nucleotide-specific Raman signals, which is also suitable for biomolecule sensing.

The integration of Artificial Intelligence (AI) techniques with medical kits has revolutionized disease diagnosis, enabling rapid and accurate identification of various conditions. We developed a novel deep learning model, namely DeepATsers based on a combination of CNN and GAN to employ a one-pot SERS biosensor to rapidly detect COVID-19 infection. The model accurately identifies each SARS-CoV-2 protein (S protein, N protein, VLP protein, Streptavidin protein, and blank signal) from its experimental fingerprint-like spectral data introduced in this study. Several augmentation techniques such as EMSA, Gaussian-noise, GAN, and K-fold cross-validation, and their combinations were utilized for the SERS spectral dataset generalization and prevented model overfitting. The original experimental dataset of 126 spectra was augmented to 780 spectra that resembled the original set by using GAN with a low KL divergence value of 0.02. This significantly improves the average accuracy of protein classification from 0.6000 to 0.9750. The deep learning model deployed optimal hyperparameters and outperformed in most measurements comparing supervised machine learning methods such as RF, GBM, SVM, and KNN, both with and without augmented spectral datasets. For model training, a whole range of spectra wavenumbers ( $$320 \hbox { cm}^{-1}$$ to $$1650 \hbox { cm}^{-1}$$ ) as well as wavenumbers ( $$1078 \hbox { cm}^{-1}$$ and $$1582 \hbox { cm}^{-1}$$ ) only for fingerprint peak spectra were employed. The former led to highly accurate 0.9750 predictions in comparison to 0.4318 for the latter one. Finally, independent experimental spectra of SARS-CoV-2 Omicron variant were used in the model verification. Thus, DeepATsers can be considered a robust, generalized, and generative deep learning framework for 1D SERS spectral datasets of SARS-CoV-2.

Synergistic drug combination prediction tasks based on the computational models have been widely studied and applied in the cancer field. However, most of models only consider the interactions between drug pairs and specific cell lines, without taking into account the multiple biological relationships of drug-drug and cell line-cell line that also largely affect synergistic mechanisms. To this end, here we propose a multi-modal deep learning framework, termed MDNNSyn, which adequately applies multi-source information and trains multi-modal features to infer potential synergistic drug combinations. MDNNSyn extracts topology modality features by implementing the multi-layer hypergraph neural network on drug synergy hypergraph and constructs semantic modality features through similarity strategy. A multi-modal fusion network layer with gated neural network is then employed for synergy score prediction. MDNNSyn is compared to five classic and state-of-the-art prediction methods on DrugCombDB and Oncology-Screen datasets. The model achieves area under the curve (AUC) scores of 0.8682 and 0.9013 on two datasets, an improvement of 3.70% and 2.71% over the second-best model. Case study indicates that MDNNSyn is capable of detecting potential synergistic drug combinations.

A fast approximation for the Euclidean distance, i.e., the \(L_2\)-norm of a complex number was investigated. The problem is obtaining an initial estimate with minimal effort and maximum numerical precision. We show how to eliminate the square root computation and improve the precision of an approximation with just a few shifts and additions from 41% to just 4%. This corresponds to an improvement of precision from 1 bit to almost 5 bits. As successive Newton iterations double the number of bits, this eliminates the need for two initial iterations. For example, a result with 16-bit precision can be obtained by just two iterations instead of four. We implement the computation by using two types of basic neurons, namely a high-pass (HP) neuron and a low-pass (LP) neuron, in the Spiking Neural P systems framework.