Precision pharmacology aims to manipulate specific cellular interactions within complex tissues. In this pursuit, we introduce DART.2 (drug acutely restricted by tethering), a second-generation cell-specific pharmacology technology. The core advance is optimized cellular specificity—up to 3,000-fold in 15 min—enabling the targeted delivery of even epileptogenic drugs without off-target effects. Additionally, we introduce brain-wide dosing methods as an alternative to local cannulation and tracer reagents for brain-wide dose quantification. We describe four pharmaceuticals—two that antagonize excitatory and inhibitory postsynaptic receptors, and two that allosterically potentiate these receptors. Their versatility is showcased across multiple mouse-brain regions, including cerebellum, striatum, visual cortex and retina. Finally, in the ventral tegmental area, we find that blocking inhibitory inputs to dopamine neurons accelerates locomotion, contrasting with previous optogenetic and pharmacological findings. Beyond enabling the bidirectional perturbation of chemical synapses, these reagents offer intersectional precision—between genetically defined postsynaptic cells and neurotransmitter-defined presynaptic partners. DART (drug acutely restricted by tethering) enables the manipulation of native receptors on genetically defined neurons. This work describes second-generation DART reagents for manipulating GABAA and AMPA receptors with higher cellular specificity than previously achieved.

Continual learning requires a model to adapt to ongoing changes in the data distribution, and often to the set of tasks to be performed. It is rare, however, that the data and task changes are completely unpredictable. Given a description of an overarching goal or data theme, which we call a realm, humans can often guess what concepts are associated with it. We show here that the combination of a large language model and an image generation model can similarly provide useful premonitions as to how a continual learning challenge might develop over time. We use the large language model to generate text descriptions of semantically related classes that might potentially appear in the data stream in future. These descriptions are then rendered using Stable Diffusion to generate new labelled image samples. The resulting synthetic dataset is employed for supervised pre-training, but is discarded prior to commencing continual learning, along with the pre-training classification head. We find that the backbone of our pre-trained networks can learn representations useful for the downstream continual learning problem, thus becoming a valuable input to any existing continual learning method. Although there are complexities arising from the domain gap between real and synthetic images, we show that pre-training models in this manner improves multiple Class Incremenal Learning (CIL) methods on fine-grained image classification benchmarks. Supporting code can be found at https://github.com/cl-premonition/premonition.

KLS-13019 was reported previously to reverse paclitaxel-induced mechanical allodynia in a mouse model of chemotherapy-induced peripheral neuropathy (CIPN). Recent studies demonstrated that paclitaxel-induced increases in inflammatory markers (GPR55, NLRP3 and IL-1b) of dorsal root ganglion (DRG) cultures were shown to be reversed by KLS-13019 treatment. The mechanism of action for KLS-13019-mediated reversal of paclitaxel-induced neuroinflammation now has been explored using GPR55 siRNA. Pretreatment of DRG cultures with GPR55 siRNA produced a 21% decrease of immunoreactive (IR) area for GPR55 in cell bodies and a 59% decrease in neuritic IR area, as determined by high content imaging. Using a 24-hour reversal treatment paradigm, paclitaxel-induced increases in the inflammatory markers were reversed back to control levels after KLS-3019 treatment. Decreases in these inflammatory markers produced by KLS-13019 were significantly attenuated by GPR55 siRNA co-treatment, with mean IR area responses being attenuated by 56% in neurites and 53% in cell bodies. These data indicate that the percentage decreases in siRNA-mediated attenuation of KLS-13019-related efficacy on the inflammatory markers were similar to the percentage knockdown observed for neuritic GPR55 IR area. Similar studies conducted with cannabidiol (CBD), the parent compound of KLS-13019, produced low efficacy (25%) reversal of all inflammatory markers that were poorly attenuated (29%) by GPR55 siRNA. CBD was shown previously to be ineffective in reversing paclitaxel-induced mechanical allodynia. The present studies indicated significant differences between the anti-inflammatory properties of KLS-13019 and CBD which may play a role in their observed differences in the reversibility of mechanical allodynia in a mouse model of CIPN.

Artificial Intelligence and Machine Learning is a highly active area of research across numerous subgenres. One such example is Reinforcement Learning, which relies on trial and error based sampling of the environment to train an agent at completing a given task, commonly applied to Atari 2600 games within research applications. Many variants of Reinforcement Learning algorithms exist, all of which apply numerous hyperparameters to control the learning process in some way, from strength of backpropagation updates through to rates of exploration. The breadth of choice across these hyperparameters makes optimal training a challenging task, with no feedback given until significant time has lapsed in training. What is of interest here is the relative importance across these hyperparameters as well as any relationship amongst them at deriving high performance agents. Through this research we apply the common statistical approach of Design of Experiments to the task of understanding the state space of the numerous hyperparameters present in Reinforcement Learning algorithms such as the Double Deep Q-Network and Prioritized Experience Replay methods. We identify the learning rate as the only primary contributor of success or failure of value based Reinforcement Learning approaches to achieve optimal reward gain. This finding suggests the possibility of significantly reduced effort and time for considering the effects of non-dependent hyperparameters.

Continual learning (CL) aims to incrementally learn different tasks (such as classification) in a non-stationary data stream without forgetting old ones. Most CL works focus on tackling catastrophic forgetting under a learning-from-scratch paradigm. However, with the increasing prominence of foundation models, pre-trained models equipped with informative representations have become available for various downstream requirements. Several CL methods based on pre-trained models have been explored, either utilizing pre-extracted features directly (which makes bridging distribution gaps challenging) or incorporating adaptors (which may be subject to forgetting). In this paper, we propose a concise and effective approach for CL with pre-trained models. Given that forgetting occurs during parameter updating, we contemplate an alternative approach that exploits training-free random projectors and class-prototype accumulation, which thus bypasses the issue. Specifically, we inject a frozen Random Projection layer with nonlinear activation between the pre-trained model's feature representations and output head, which captures interactions between features with expanded dimensionality, providing enhanced linear separability for class-prototype-based CL. We also demonstrate the importance of decorrelating the class-prototypes to reduce the distribution disparity when using pre-trained representations. These techniques prove to be effective and circumvent the problem of forgetting for both class- and domain-incremental continual learning. Compared to previous methods applied to pre-trained ViT-B/16 models, we reduce final error rates by between 20% and 62% on seven class-incremental benchmark datasets, despite not using any rehearsal memory. We conclude that the full potential of pre-trained models for simple, effective, and fast continual learning has not hitherto been fully tapped. Code is available at https://github.com/RanPAC/RanPAC.

Hydraulic conductivity is one of the most challenging hydrogeological properties to appropriately measure due to its dependence on the measurement scale and the influence of heterogeneity. This paper presents a comparison of saturated hydraulic conductivities ( K ) determined for a quasi-homogeneous coastal sand aquifer, estimated using eight different methodologies, encompassing empirical, hydraulic and numerical modeling methods. The geometric means of K, determined using 22 methods, spanning measurement scales varying between 0.01 and 100 m, ranged between 3.6 and 58.3 m/d. K estimates from Cone Penetration Test (CPT) data proved wider than those obtained using the other methods, while various empirical equations, commonly used to estimate K from grain-size analysis and Tide-Aquifer interaction techniques revealed variations of up to one order of magnitude. Single-well tracer dilution tests provided an alternative for making preliminary estimates of K when hydraulic gradients were known. Estimates from the slug tests proved between 1.2 and 1.6 times larger than those determined from pumping tests which, with one of the smallest ranges of variation, provided a representative average K of the aquifer as revealed by numerical modeling. By contrast, variations in K with depth could be detected at small scales (~ 0.1 m). Hydraulic Profiling Tool (HPT) system data indicated that K decreases with depth, which was supported by the numerical model results. No scale effect on K was apparent when considering the ensemble of results, suggesting that hydraulic conductivity estimates do not depend on the scale of measurement in the absence of significant aquifer heterogeneities.

Conventional views of saltwater intrusion (SWI), where a basal saline wedge extends inland below fresh groundwater, can be complicated by the influence of saltwater cells in the upper part of aquifers in areas affected by tidal cycles. Distinguishing the contribution of each saltwater source may prove fundamental for well design and resource management. Application of time-lapse electrical resistivity imaging (ERI) during a 32-h pumping test in a pristine unconfined coastal sand aquifer, affected by strong tidal ranges (>2 m), aimed to evaluate the potential of the method to characterize the source of induced SWI in four dimensions (three dimensions and time). Water level monitoring during the test revealed that at the end of pumping, the upper 2 m of the aquifer had dewatered in the vicinity of the well field, reversing hydraulic gradients between the aquifer and the sea. This induced SI, with mixing models of well head water quality suggesting that saline water contributions to total discharge rose from 4 % to 8 %. ERI results reflected dewatering through an increase in resistivity in the upper 2-6 m of the aquifer, while a decline in resistivity, relative to background conditions, occurred immediately below this, reflecting the migration of saline water through the upper layers of the aquifer to the pumping well. By contrast no change in resistivity occurred at depth, indicating no significant change in contribution from the basal saline water to discharge. Test findings suggest that future water resource development at the site should focus on close monitoring of shallow pumping, or pumping from deeper parts of the aquifer, while more generally demonstrating the value of time-lapse geophysical methods in informing coastal water resource management.

Dark deposits visible from orbit appear in the Martian south polar region during the springtime. These are thought to form from explosive jets of carbon dioxide gas breaking through the thawing seasonal ice cap, carrying dust and dirt which is then deposited onto the ice as dark 'blotches', or blown by the surface winds into streaks or 'fans'. We investigate machine learning (ML) methods for automatically identifying these seasonal features in High Resolution Imaging Science Experiment (HiRISE) satellite imagery. We designed deep Convolutional Neural Networks (CNNs) that were trained and tested using the catalog generated by Planet Four, an online citizen science project mapping the south polar seasonal deposits. We validated the CNNs by comparing their results with those of ISODATA (Iterative Self-Organizing Data Analysis Technique) clustering and as expected, the CNNs were significantly better at predicting the results found by Planet Four, in both the area of predicted seasonal deposits and in delineating their boundaries. We found neither the CNNs or ISODATA were suited to predicting the source point and directions of seasonal fans, which is a strength of the citizen science approach. The CNNs showed good agreement with Planet Four in cross-validation metrics and detected some seasonal deposits in the HiRISE images missed in the Planet Four catalog; the total area of seasonal deposits predicted by the CNNs was 27% larger than that of the Planet Four catalog, but this aspect varied considerably on a per-image basis.

KLS-13019 is a structural analogue of cannabidiol, that shows improved bioavailability and potency in both preventing and reversing paclitaxel-induced neurotoxicity in vitro and in vivo. KLS-13019 was selected as a development candidate and attention was turned to development of a scalable synthesis. The original synthesis of KLS-13019 was not efficient, regioselective, or high yielding. Two new syntheses are reported that make use of the palladium catalyzed cross couplings to a chemically advanced intermediate 5, dramatically shortening (3-4 steps) and improving the overall yield. In addition, a convenient one pot Boc-cleavage and acetylation procedure is described to avoid impurities generated from a step-wise process.

Tidal forcing influences groundwater flow and salt distribution in shallow coastal aquifers, with the interaction between sea level variations and geology proving fundamental for assessing the risk of seawater intrusion (SI). Constraining the relative importance of each is often confounded by the influences of groundwater abstraction and geological heterogeneity, with understanding of the latter often restricted by sampling point availability and poor spatial resolution. This paper describes the application of geophysical and geotechnical methods to better characterize groundwater salinity patterns in a tidally dominated ~ 20 m thick sequence of beach sand, unaffected by groundwater abstraction. Electrical resistivity tomography (ERT) revealed the deposit to consist of an upper wedge of low resistivity (< 3 Ωm), reaching over 8 m thick in the vicinity of the low water mark, overlying a higher resistivity unit. Cone penetrometer testing (CPT), and associated high-resolution hydraulic profiling tool system (HPT), coupled with water quality sampling, revealed the wedge to reflect an intertidal recirculation cell (IRC), which restricts freshwater discharge from a relatively homogeneous sand unit to a zone of seepage within the IRC. The application of CPT and HPT techniques underscored the value of geotechnical methods in distinguishing between geological and water quality contributions to geophysical responses. Survey results have permitted a clear characterization of the groundwater flow regime in a coastal aquifer with an IRC, highlighting the benefit of combining geophysical and geotechnical methods to better characterize shallow SI mechanisms and groundwater flow in coastal hydrogeological environments.

ABSTRACT We investigate the effect of personal budget management services provided by a financial institution on undesirable debts of 4,256 individuals and families over a period of one year using objective budget data spanning three years. Results from this study using an Australian dataset show that the majority of the participants reduced their debt repayments and debt repayment to income ratio over a period of one year, Our results also show that the majority of the participants hod an increase in funds available for living expenses and savings.

Investigation of value based Reinforcement Learning algorithms exhibited a resurgence into mainstream research in 2015 following demonstration of super-human performance when applied to Atari 2600 games. Since then, significant media attention and hype have accompanied this area, and the field of Artificial Intelligence generally, spread across distinct categories. This review paper is focused exclusively on the progression of value based Reinforcement Learning in the last five years. We aim to distill the incremental improvements to stability and performance in this period, highlighting the minimal changes to the base algorithm over this time. This holds true with all but the one exception of the Recurrent Experience Replay in Distributed Reinforcement Learning algorithm, representing a fundamental shift and increase in agent performance through an advanced memory representation. We suggest a new focus area for value based Reinforcement Learning research.

SUMMARYCell-specific pharmaceutical technologies promise mechanistic insight into clinical drugs―those that treat, and often define, human disease. In particular,DART(drug acutely restricted by tethering) achieves genetically programmable control of drug concentration over cellular dimensions. The method is compatible with clinical pharmaceuticals and amenable to studies in behaving animals. Here, we describeDART.2, comprising three advances. First, we improve the efficiency of chemical capture, enabling cell-specific accumulation of drug to ∼3,000-times the ambient concentration in 15 min. Second, we develop tracer reagents, providing a behavior-independent measure of cellular target engagement in each animal. Third, we extend the method to positive allosteric modulators and outline design principles for this clinically significant class. We showcase the platform with four pharmaceuticals―two that weaken excitatory (AMPAR) or inhibitory (GABAAR) chemical neurotransmission, and two that strengthen these forms of synaptic communication. Across four labs, we tested reagents in the mouse cerebellum, basal ganglia, visual cortex, and retina. Collectively, we demonstrate robust, bidirectional editing of chemical neurotransmission. We provide for distribution of validated reagents, community design principles, and synthetic building blocks for application to diverse pharmaceuticals.

KLS-13019, a novel devised cannabinoid-like compound, was explored for anti-inflammatory actions in dorsal root ganglion cultures relevant to chemotherapy-induced peripheral neuropathy (CIPN). Time course studies with 3 µM paclitaxel indicated > 1.9-fold increases in immunoreactive (IR) area for cell body GPR55 after 30 min as determined by high content imaging. To test for reversibility of paclitaxel-induced increases in GPR55, cultures were treated for 8 h with paclitaxel alone and then a dose response to KLS-13019 added for another 16 h. This “reversal” paradigm indicated established increases in cell body GPR55 IR areas were decreased back to control levels. Because GPR55 had previously reported inflammatory actions, IL-1β and NLRP3 (inflammasome-3 marker) were also measured in the “reversal” paradigm. Significant increases in all inflammatory markers were produced after 8 h of paclitaxel treatment alone that were reversed to control levels with KLS-13019 treatment. Accompanying studies using alamar blue indicated that decreased cellular viability produced by paclitaxel treatment was reverted back to control levels by KLS-13019. Similar studies conducted with lysophosphatidylinositol (GPR55 agonist) in DRG or hippocampal cultures demonstrated significant increases in neuritic GPR55, NLRP3 and IL-1β areas that were reversed to control levels with KLS-13019 treatment. Studies with a human GPR55-β-arrestin assay in Discover X cells indicated that KLS-13019 was an antagonist without agonist activity. These studies indicated that KLS-13019 has anti-inflammatory properties mediated through GPR55 antagonist actions. Together with previous studies, KLS-13019 is a potent neuroprotective, anti-inflammatory cannabinoid with therapeutic potential for high efficacy treatment of neuropathic pain.

Matrix assisted laser desorption/ionization mass spectrometry imaging (MALDI MSI) can determine the spatial distribution of analytes such as protein distributions in a tissue section according to their mass-to-charge ratio. Here, we explored the clinical potential of machine learning (ML) applied to MALDI MSI data for cancer diagnostic classification using tissue microarrays (TMAs) on 302 colorectal (CRC) and 257 endometrial cancer (EC)) patients. ML based on deep neural networks discriminated colorectal tumour from normal tissue with an overall accuracy of 98% in balanced cross-validation (98.2% sensitivity and 98.6% specificity). Moreover, our machine learning approach predicted the presence of lymph node metastasis (LNM) for primary tumours of EC with an accuracy of 80% (90% sensitivity and 69% specificity). Our results demonstrate the capability of MALDI MSI for complementing classic histopathological examination for cancer diagnostic applications.

Despite the development and success of cochlear implants over several decades, wide inter-subject variability in speech perception is reported. This suggests that cochlear implant user-dependent factors limit speech perception at the individual level. Clinical studies have demonstrated the importance of the number, placement, and insertion depths of electrodes on speech recognition abilities. However, these do not account for all inter-subject variability and to what extent these factors affect speech recognition abilities has not been studied. In this paper, an information theoretic method and machine learning technique are unified in a model to investigate the extent to which key factors limit cochlear implant electrode discrimination. The framework uses a neural network classifier to predict which electrode is stimulated for a given simulated activation pattern of the auditory nerve, and mutual information is then estimated between the actual stimulated electrode and predicted ones. We also investigate how and to what extent the choices of parameters affect the performance of the model. The advantages of this framework include i) electrode discrimination ability is quantified using information theory, ii) it provides a flexible framework that may be used to investigate the key factors that limit the performance of cochlear implant users, and iii) it provides insights for future modeling studies of other types of neural prostheses.

Laboratory experiments and numerical simulations were utilized in this study to assess the impact of aquifer stratification on saltwater intrusion. Three homogeneous and six layered aquifers were investigated. Image processing algorithms facilitated the precise calculation of saltwater wedge toe length, width of the mixing zone, and angle of intrusion. It was concluded that the length of intrusion in stratified aquifers is predominantly a function of permeability contrast, total aquifer transmissivity and the number of heterogeneous layers, being positively correlated to all three. When a lower permeability layer overlays or underlays more permeable zones its mixing zone widens, while it becomes thinner for the higher permeability strata. The change in the width of the mixing zone (WMZ) is positively correlated to permeability contrast, while it applies to all strata irrespectively of their relative vertical position in the aquifer. Variations in the applied hydraulic head causes the transient widening of WMZ. These peak WMZ values are larger during saltwater retreat and are negatively correlated to the layer's permeability and distance from the aquifer's bottom. Moreover, steeper angles of intrusion are observed in cases where low permeability layers overlay more permeable strata, and milder ones in the inverse aquifer setups. The presence of a low permeability upper layer results in the confinement of the saltwater wedge in the lower part of the stratified aquifer. This occurs until a critical hydraulic head difference is applied to the system. This hydraulic gradient value was found to be a function of layer width and permeability contrast alike.

Background: Machine learning (ML) can harness information from large databases with complex structures. We present a simple and fast hypothesis-free machine learning pipeline for risk factor discovery that accounts for non-linearity and interaction in large biomedical databases with minimal variable pre-processing. Methods: Mortality models were built using gradient boosting decision trees (GBDT) and important predictors were identified using SHAP values. Cox models controlled for false discovery rate were used for interpretability and further validation. The pipeline was tested using information from 502,506 UK Biobank participants, aged 37 to 73 years at recruitment and followed over seven years for mortality registrations. Results: From the 11,639 predictors included in GBDT, 193 potential risk factors had SHAP values 0.05 or greater and were selected for further modelling. Of the total variable importance summed up, 60% was directly health related, and baseline characteristics, sociodemographics, and lifestyle factors each contributed about 10%. Cox models adjusted for baseline characteristics, showed evidence for an association with mortality for 166 out of the 193 predictors. For 19 predictors we saw evidence for an association in the unadjusted but not adjusted analyses, suggesting confounding by basic characteristics. Identified "important" predictors included traditional risk factors such as age, sex, ethnicity, education, material deprivation, smoking, physical activity, self-rated health, BMI, hypertension, cardio-vascular diseases, cancer diagnoses and type 2 diabetes, as confirmed by previous studies. Conclusion: Our approach provides a fast and pragmatic solution for hypothesis free risk factor identification.

Background Glioblastoma is the most aggressive type of brain cancer with high-levels of intra- and inter-tumour heterogeneity that contribute to its rapid growth and invasion within the brain. However, a spatial characterisation of gene signatures and the cell types expressing these in different tumour locations is still lacking. Methods We have used a deep convolutional neural network (DCNN) as a semantic segmentation model to segment seven different tumour regions including leading edge (LE), infiltrating tumour (IT), cellular tumour (CT), cellular tumour microvascular proliferation (CTmvp), cellular tumour pseudopalisading region around necrosis (CTpan), cellular tumour perinecrotic zones (CTpnz) and cellular tumour necrosis (CTne) in digitised glioblastoma histopathological slides from The Cancer Genome Atlas (TCGA). Correlation analysis between segmentation results from tumour images together with matched RNA expression data was performed to identify genetic signatures that are specific to different tumour regions. Results We found that spatially resolved gene signatures were strongly correlated with survival in patients with defined genetic mutations. Further in silico cell ontology analysis along with single-cell RNA sequencing data from resected glioblastoma tissue samples showed that these tumour regions had different gene signatures, whose expression was driven by different cell types in the regional tumour microenvironment. Our results further pointed to a key role for interactions between microglia/pericytes/monocytes and tumour cells that occur in the IT and CTmvp regions, which may contribute to poor patient survival. Conclusions This work identified key histopathological features that correlate with patient survival and detected spatially associated genetic signatures that contribute to tumour-stroma interactions and which should be investigated as new targets in glioblastoma. The source codes and datasets used are available in GitHub: https://github.com/amin20/GBM_WSSM .