Bayesian inference for identifying tumour-specific cancer dependencies through integration of ex-vivo drug response assays and drug-protein profiling.

The identification of tumor-specific molecular dependencies is essential for the development of effective cancer therapies. Genetic and chemical perturbations are powerful tools for discovering these dependencies. Even though chemical perturbations can be applied to primary cancer samples at large scale, the interpretation of experiment outcomes is often complicated by the fact that one chemical compound can affect multiple proteins. To overcome this challenge, Batzilla et al. (PLoS Comput Biol 18(8): e1010438, 2022) proposed DepInfeR, a regularized multi-response regression model designed to identify and estimate specific molecular dependencies of individual cancers from their ex-vivo drug sensitivity profiles. Inspired by their work, we propose a Bayesian extension to DepInfeR. Our proposed approach offers several advantages over DepInfeR, including e.g. the ability to handle missing values in both protein-drug affinity and drug sensitivity profiles without the need for data pre-processing steps such as imputation. Moreover, our approach uses Gaussian Processes to capture more complex molecular dependency structures, and provides probabilistic statements about whether a protein in the protein-drug affinity profiles is informative to the drug sensitivity profiles. Simulation studies demonstrate that our proposed approach achieves better prediction accuracy, and is able to discover unreported dependency structures.

Facial Micro-Expression Recognition Using Double-Stream 3D Convolutional Neural Network with Domain Adaptation.

Humans show micro-expressions (MEs) under some circumstances. MEs are a display of emotions that a human wants to conceal. The recognition of MEs has been applied in various fields. However, automatic ME recognition remains a challenging problem due to two major obstacles. As MEs are typically of short duration and low intensity, it is hard to extract discriminative features from ME videos. Moreover, it is tedious to collect ME data. Existing ME datasets usually contain insufficient video samples. In this paper, we propose a deep learning model, double-stream 3D convolutional neural network (DS-3DCNN), for recognizing MEs captured in video. The recognition framework contains two streams of 3D-CNN. The first extracts spatiotemporal features from the raw ME videos. The second extracts variations of the facial motions within the spatiotemporal domain. To facilitate feature extraction, the subtle motion embedded in a ME is amplified. To address the insufficient ME data, a macro-expression dataset is employed to expand the training sample size. Supervised domain adaptation is adopted in model training in order to bridge the difference between ME and macro-expression datasets. The DS-3DCNN model is evaluated on two publicly available ME datasets. The results show that the model outperforms various state-of-the-art models; in particular, the model outperformed the best model presented in MEGC2019 by more than 6%.

Application of Micro-Engineered Kidney, Liver, and Respiratory System Models to Accelerate Preclinical Drug Testing and Development.

Developing novel drug formulations and progressing them to the clinical environment relies on preclinical in vitro studies and animal tests to evaluate efficacy and toxicity. However, these current techniques have failed to accurately predict the clinical success of new therapies with a high degree of certainty. The main reason for this failure is that conventional in vitro tissue models lack numerous physiological characteristics of human organs, such as biomechanical forces and biofluid flow. Moreover, animal models often fail to recapitulate the physiology, anatomy, and mechanisms of disease development in human. These shortfalls often lead to failure in drug development, with substantial time and money spent. To tackle this issue, organ-on-chip technology offers realistic in vitro human organ models that mimic the physiology of tissues, including biomechanical forces, stress, strain, cellular heterogeneity, and the interaction between multiple tissues and their simultaneous responses to a therapy. For the latter, complex networks of multiple-organ models are constructed together, known as multiple-organs-on-chip. Numerous studies have demonstrated successful application of organ-on-chips for drug testing, with results comparable to clinical outcomes. This review will summarize and critically evaluate these studies, with a focus on kidney, liver, and respiratory system-on-chip models, and will discuss their progress in their application as a preclinical drug-testing platform to determine in vitro drug toxicology, metabolism, and transport. Further, the advances in the design of these models for improving preclinical drug testing as well as the opportunities for future work will be discussed.

Magnetic Imprinted Polymer-Based Quartz Crystal Microbalance Sensor for Sensitive Label-Free Detection of Methylene Blue in Groundwater.

Tiny changes in the mass of the sensor in a quartz crystal microbalance with dissipation monitoring (QCM-D) can be observed. However, the lack of specificity for target species has hindered the use of QCM-D. Here, molecularly imprinted polymers (MIPs) were used to modify a QCM-D sensor to provide specificity. The MIPs were formed in the presence of sodium dodecyl benzene sulfonate. Imprinted layers on Fe3O4 nanoparticles were formed using pyrrole as the functional monomer and cross-linker and methylene blue (MB) as a template. The MIPs produced were then attached to the surface of a QCM-D sensor. The MIPs-coated QCM-D sensor could recognize MB and gave a linear response in the concentration range 25 to 1.5 × 102 µg/L and a detection limit of 1.4 µg/L. The QCM-D sensor was selective for MB over structural analogs. The MIPs-coated QCM-D sensor was successfully used to detect MB in river water and seawater samples, and the recoveries were good. This is the first time MB has been detected using a QCM-D sensor. Mass is an intrinsic property of matter, so this method could easily be extended to other target species by using different MIPs.

Optimization, chain conformation and characterization of exopolysaccharide isolated from Leuconostoc mesenteroides DRP105.

Response surface methodology (RSM) was used to optimize the fermentation condition of exopolysaccharide (EPS) producing strain Leuconostoc mesenteroides DRP105. Result showed that the optimum condition was sucrose 86.83g/L, tryptone 15.47g/L, initial pH7.18 and maximum yield was 53.79±0.78g/L in 36h fermentation. Chain conformation was characterized by Congo red test, ß-elimination and circular dichroism (CD), which indicated that the EPS was O-linkage and exhibited random coil structure in aqueous solution. CD results concluded hydrogen-bond interaction and chirality might be connected with concentration. Purified EPS has a higher degradation temperature of 279.42°C, suggesting high thermal stability of the EPS. The absolute value of zeta potential and particle size were enhanced with increasing concentration. Crude EPS and its purified fraction were found to have moderate DPPH, hydroxyl, superoxide anion radicals scavenging activities and reducing power. This study provided a high yield EPS with unique characteristics for industrial applications.

Optimization, purification and structural characterization of a dextran produced by L. mesenteroides isolated from Chinese sauerkraut.

A higher yield of dextran strain Leuconstoc mesenteroides TDS2-19 was isolated from Chinese sauerkraut juice. Effects of the three main factors on exopolysaccharide (EPS) yield were investigated by central composite design (CCD) and the optimum composition was sucrose 117.48g/L, sodium acetate 4.10g/L, and initial pH 6.88. Optimum results showed that EPS yield was increased to 71.23±2.25g/L in 48h fermentation, 31.24% higher than before. The molecular weight (Mw) of ESP was 8.79×107Da, as determined by high-performance size-exclusion chromatography (HPSEC). Fourier transform infrared spectra (FT-IR) and nuclear magnetic resonance spectra (NMR) showed that the polysaccharide synthesized by L. mesenteroides TDS2-19 in the MRS medium was dextran with a peak, a linear backbone composed of consecutive α-(1→6)-linked d-glucopyranose units. No branching was observed in the dextran structure. The present study suggested that L. mesenteroides TDS2-19 might be used for the industrial-scale production of linear dextran.