Deep learning for single-molecule science

Single-Molecule Bioelectronic Sensors with AI-Aided Data Analysis: Convergence and Challenges

Single-molecule bioelectronic sensing, a groundbreaking domain in biological research, has revolutionized our understanding of molecules by revealing deep insights into fundamental biological processes. The advent of emergent technologies, such as nanogapped electrodes and nanopores, has greatly enhanced this field, providing exceptional sensitivity, resolution, and integration capabilities. However, challenges persist, such as complex data sets with high noise levels and stochastic molecular dynamics. Artificial intelligence (AI) has stepped in to address these issues with its powerful data processing capabilities. AI algorithms effectively extract meaningful features, detect subtle changes, improve signal-to-noise ratios, and uncover hidden patterns in massive data. This review explores the synergy between AI and single-molecule bioelectronic sensing, focusing on how AI enhances signal processing and data analysis to boost accuracy and reliability. We also discuss current limitations and future directions for integrating AI, highlighting its potential to advance biological research and technological innovation.

Machine learning empowered next generation DNA sequencing: perspective and prospectus

The pursuit of ultra-rapid, cost-effective, and accurate DNA sequencing is a highly sought after aspect of personalized medicine development. With recent advancements, mainstream machine learning (ML) algorithms hold immense promise for high throughput DNA sequencing at the single nucleotide level. While ML has revolutionized multiple domains of nanoscience and nanotechnology, its implementation in DNA sequencing is still in its preliminary stages. ML-aided DNA sequencing is especially appealing, as ML has the potential to decipher complex patterns and extract knowledge from complex datasets. Herein, we present a holistic framework of ML-aided next-generation DNA sequencing with domain knowledge to set directions toward the development of artificially intelligent DNA sequencers. This perspective focuses on the current state-of-the-art ML-aided DNA sequencing, exploring the opportunities as well as the future challenges in this field. In addition, we provide our personal viewpoints on the critical issues that require attention in the context of ML-aided DNA sequencing.

Offloading the computational complexity of transfer learning with generic features

Deep learning approaches are generally complex, requiring extensive computational resources and having high time complexity. Transfer learning is a state-of-the-art approach to reducing the requirements of high computational resources by using pre-trained models without compromising accuracy and performance. In conventional studies, pre-trained models are trained on datasets from different but similar domains with many domain-specific features. The computational requirements of transfer learning are directly dependent on the number of features that include the domain-specific and the generic features. This article investigates the prospects of reducing the computational requirements of the transfer learning models by discarding domain-specific features from a pre-trained model. The approach is applied to breast cancer detection using the dataset curated breast imaging subset of the digital database for screening mammography and various performance metrics such as precision, accuracy, recall, F1-score, and computational requirements. It is seen that discarding the domain-specific features to a specific limit provides significant performance improvements as well as minimizes the computational requirements in terms of training time (reduced by approx. 12%), processor utilization (reduced approx. 25%), and memory usage (reduced approx. 22%). The proposed transfer learning strategy increases accuracy (approx. 7%) and offloads computational complexity expeditiously.

Total variation denoising-based method of identifying the states of single molecules in break junction data

Break junction (BJ) measurements provide insights into the electrical properties of diverse molecules, enabling the direct assessment of single-molecule conductances. The BJ method displays potential for use in determining the dynamics of individual molecules, single-molecule chemical reactions, and biomolecules, such as deoxyribonucleic acid and ribonucleic acid. However, conductance data obtained via single-molecule measurements may be susceptible to fluctuations due to minute structural changes within the junctions. Consequently, clearly identifying the conduction states of these molecules is challenging. This study aims to develop a method of precisely identifying conduction state traces. We propose a novel single-molecule analysis approach that employs total variation denoising (TVD) in signal processing, focusing on the integration of information technology with measured single-molecule data. We successfully applied this method to simulated conductance traces, effectively denoise the data, and elucidate multiple conduction states. The proposed method facilitates the identification of well-defined plateau lengths and supervised machine learning with enhanced accuracies. The introduced TVD-based analytical method is effective in elucidating the states within the measured single-molecule data. This approach exhibits the potential to offer novel perspectives regarding the formation of molecular junctions, conformational changes, and cleavage.

A perspective on computer vision in biosensing

Computer vision has become a powerful tool in the field of biosensing, aiding in the development of innovative and precise systems for the analysis and interpretation of biological data. This interdisciplinary approach harnesses the capabilities of computer vision algorithms and techniques to extract valuable information from various biosensing applications, including medical diagnostics, environmental monitoring, and food health. Despite years of development, there is still significant room for improvement in this area. In this perspective, we outline how computer vision is applied to raw sensor data in biosensors and its advantages to biosensing applications. We then discuss ongoing research and developments in the field and subsequently explore the challenges and opportunities that computer vision faces in biosensor applications. We also suggest directions for future work, ultimately underscoring the significant impact of computer vision on advancing biosensing technologies and their applications.

What Is Machine Learning, Artificial Neural Networks and Deep Learning?-Examples of Practical Applications in Medicine

Machine learning (ML), artificial neural networks (ANNs), and deep learning (DL) are all topics that fall under the heading of artificial intelligence (AI) and have gained popularity in recent years. ML involves the application of algorithms to automate decision-making processes using models that have not been manually programmed but have been trained on data. ANNs that are a part of ML aim to simulate the structure and function of the human brain. DL, on the other hand, uses multiple layers of interconnected neurons. This enables the processing and analysis of large and complex databases. In medicine, these techniques are being introduced to improve the speed and efficiency of disease diagnosis and treatment. Each of the AI techniques presented in the paper is supported with an example of a possible medical application. Given the rapid development of technology, the use of AI in medicine shows promising results in the context of patient care. It is particularly important to keep a close eye on this issue and conduct further research in order to fully explore the potential of ML, ANNs, and DL, and bring further applications into clinical use in the future.

Unsupervised classification of voltammetric data beyond principal component analysis

In this study, we evaluate different apoproaches to unsupervised classification of cyclic voltammetric data, including Principal Component Analysis (PCA), t-distributed Stochastic Neighbour Embedding (t-SNE), Uniform Manifold Approximation and Projection (UMAP) as well as neural networks. To this end, we exploit a form of transfer learning, based on feature extraction in an image recognition network, VGG-16, in combination with PCA, t-SNE or UMAP. Overall, we find that t-SNE performs best when applied directly to numerical data (noise-free case) or to features (in the presence of noise), followed by UMAP and then PCA.

Trusting our machines: validating machine learning models for single-molecule transport experiments

In this tutorial review, we will describe crucial aspects related to the application of machine learning to help users avoid the most common pitfalls. The examples we present will be based on data from the field of molecular electronics, specifically single-molecule electron transport experiments, but the concepts and problems we explore will be sufficiently general for application in other fields with similar data. In the first part of the tutorial review, we will introduce the field of single-molecule transport, and provide an overview of the most common machine learning algorithms employed. In the second part of the tutorial review, we will show, through examples grounded in single-molecule transport, that the promises of machine learning can only be fulfilled by careful application. We will end the tutorial review with a discussion of where we, as a field, could go from here.

Rational Design of Field-Effect Sensors Using Partial Differential Equations, Bayesian Inversion, and Artificial Neural Networks

Silicon nanowire field-effect transistors are promising devices used to detect minute amounts of different biological species. We introduce the theoretical and computational aspects of forward and backward modeling of biosensitive sensors. Firstly, we introduce a forward system of partial differential equations to model the electrical behavior, and secondly, a backward Bayesian Markov-chain Monte-Carlo method is used to identify the unknown parameters such as the concentration of target molecules. Furthermore, we introduce a machine learning algorithm according to multilayer feed-forward neural networks. The trained model makes it possible to predict the sensor behavior based on the given parameters.

Spiers Memorial Lecture. Next generation nanoelectrochemistry: the fundamental advances needed for applications

Nanoelectrochemistry, where electrochemical processes are controlled and investigated with nanoscale resolution, is gaining more and more attention because of the many potential applications in energy and sensing and the fact that there is much to learn about fundamental electrochemical processes when we explore them at the nanoscale. The development of instrumental methods that can explore the heterogeneity of electrochemistry occurring across an electrode surface, monitoring single molecules or many single nanoparticles on a surface simultaneously, have been pivotal in giving us new insights into nanoscale electrochemistry. Equally important has been the ability to synthesise or fabricate nanoscale entities with a high degree of control that allows us to develop nanoscale devices. Central to the latter has been the incredible advances in nanomaterial synthesis where electrode materials with atomic control over electrochemically active sites can be achieved. After introducing nanoelectrochemistry, this paper focuses on recent developments in two major application areas of nanoelectrochemistry; electrocatalysis and using single entities in sensing. Discussion of the developments in these two application fields highlights some of the advances in the fundamental understanding of nanoelectrochemical systems really driving these applications forward. Looking into our nanocrystal ball, this paper then highlights: the need to understand the impact of nanoconfinement on electrochemical processes, the need to measure many single entities, the need to develop more sophisticated ways of treating the potentially large data sets from measuring such many single entities, the need for more new methods for characterising nanoelectrochemical systems as they operate and the need for material synthesis to become more reproducible as well as possess more nanoscale control.

Deep Learning-Enabled Technologies for Bioimage Analysis

Deep learning (DL) is a subfield of machine learning (ML), which has recently demonstrated its potency to significantly improve the quantification and classification workflows in biomedical and clinical applications. Among the end applications profoundly benefitting from DL, cellular morphology quantification is one of the pioneers. Here, we first briefly explain fundamental concepts in DL and then we review some of the emerging DL-enabled applications in cell morphology quantification in the fields of embryology, point-of-care ovulation testing, as a predictive tool for fetal heart pregnancy, cancer diagnostics via classification of cancer histology images, autosomal polycystic kidney disease, and chronic kidney diseases.

A Guide to Signal Processing Algorithms for Nanopore Sensors

Nanopore technology holds great promise for a wide range of applications such as biomedical sensing, chemical detection, desalination, and energy conversion. For sensing performed in electrolytes in particular, abundant information about the translocating analytes is hidden in the fluctuating monitoring ionic current contributed from interactions between the analytes and the nanopore. Such ionic currents are inevitably affected by noise; hence, signal processing is an inseparable component of sensing in order to identify the hidden features in the signals and to analyze them. This Guide starts from untangling the signal processing flow and categorizing the various algorithms developed to extracting the useful information. By sorting the algorithms under Machine Learning (ML)-based versus non-ML-based, their underlying architectures and properties are systematically evaluated. For each category, the development tactics and features of the algorithms with implementation examples are discussed by referring to their common signal processing flow graphically summarized in a chart and by highlighting their key issues tabulated for clear comparison. How to get started with building up an ML-based algorithm is subsequently presented. The specific properties of the ML-based algorithms are then discussed in terms of learning strategy, performance evaluation, experimental repeatability and reliability, data preparation, and data utilization strategy. This Guide is concluded by outlining strategies and considerations for prospect algorithms.

Using Weakly Supervised Deep Learning to Classify and Segment Single-Molecule Break-Junction Conductance Traces

In order to design molecular electronic devices with high performance and stability, it is crucial to understand their structure-to-property relationships. Single-molecule break junction measurements yield a large number of conductance-distance traces, which are inherently highly stochastic. Here we propose a weakly supervised deep learning algorithm to classify and segment these conductance traces, a method that is mainly based on transfer learning with the pretrain-finetune technique. By exploiting the powerful feature extraction capabilities of the pretrained VGG-16 network, our convolutional neural network model not only achieves high accuracy in the classification of the conductance traces, but also segments precisely the conductance plateau from an entire trace with very few manually labeled traces. Thus, we can produce a more reliable estimation of the junction conductance and quantify the junction stability. These findings show that our model has achieved a better accuracy-to-manpower efficiency balance, opening up the possibility of using weakly supervised deep learning approaches in the studies of single-molecule junctions.

Deep Learning for Biospectroscopy and Biospectral Imaging: State-of-the-Art and Perspectives

With the advances in instrumentation and sampling techniques, there is an explosive growth of data from molecular and cellular samples. The call to extract more information from the large data sets has greatly challenged the conventional chemometrics method. Deep learning, which utilizes very large data sets for finding hidden features therein and for making accurate predictions for a wide range of applications, has been applied in an unbelievable pace in biospectroscopy and biospectral imaging in the recent 3 years. In this Feature, we first introduce the background and basic knowledge of deep learning. We then focus on the emerging applications of deep learning in the data preprocessing, feature detection, and modeling of the biological samples for spectral analysis and spectroscopic imaging. Finally, we highlight the challenges and limitations in deep learning and the outlook for future directions.

Nanostructures for Biosensing, with a Brief Overview on Cancer Detection, IoT, and the Role of Machine Learning in Smart Biosensors

Biosensors are essential tools which have been traditionally used to monitor environmental pollution and detect the presence of toxic elements and biohazardous bacteria or virus in organic matter and biomolecules for clinical diagnostics. In the last couple of decades, the scientific community has witnessed their widespread application in the fields of military, health care, industrial process control, environmental monitoring, food-quality control, and microbiology. Biosensor technology has greatly evolved from in vitro studies based on the biosensing ability of organic beings to the highly sophisticated world of nanofabrication-enabled miniaturized biosensors. The incorporation of nanotechnology in the vast field of biosensing has led to the development of novel sensors and sensing mechanisms, as well as an increase in the sensitivity and performance of the existing biosensors. Additionally, the nanoscale dimension further assists the development of sensors for rapid and simple detection in vivo as well as the ability to probe single biomolecules and obtain critical information for their detection and analysis. However, the major drawbacks of this include, but are not limited to, potential toxicities associated with the unavoidable release of nanoparticles into the environment, miniaturization-induced unreliability, lack of automation, and difficulty of integrating the nanostructured-based biosensors, as well as unreliable transduction signals from these devices. Although the field of biosensors is vast, we intend to explore various nanotechnology-enabled biosensors as part of this review article and provide a brief description of their fundamental working principles and potential applications. The article aims to provide the reader a holistic overview of different nanostructures which have been used for biosensing purposes along with some specific applications in the field of cancer detection and the Internet of things (IoT), as well as a brief overview of machine-learning-based biosensing.

Component identification and defect detection in transmission lines based on deep learning

Ensuring the stable and safe operation of the power system is an important work of the national power grid companies. The power grid company has established a special power inspection department to troubleshoot transmission line components and replace faulty components in a timely manner. At present, assisted manual inspection by drone inspection has become a trend of power line inspection. Automatically identifying component failures from images of UAV aerial transmission lines is a cutting-edge cross-cutting issue. Based on the above problems, the purpose of this article is to study the component identification and defect detection of transmission lines based on deep learning. This paper expands the dataset by adjusting the size of the convolution kernel of the CNN model and the rotation transformation of the image. The experimental results show that both methods can effectively improve the effectiveness and reliability of component identification and defect detection in transmission line inspection. The recognition and classification experiments were performed using the images collected by the drone. The experimental results show that the effectiveness and reliability of the deep learning method in the identification and defect detection of high-voltage transmission line components are very high. Faster R-CNN performs component identification and defect detection. The detection can reach a recognition speed of nearly 0.17 s per sheet, the recognition rate of the pressure-equalizing ring can reach 96.8%, and the mAP can reach 93.72%.

Plasmonic Assemblies for Real-Time Single-Molecule Biosensing

Their tunable optical properties and versatile surface functionalization have sparked applications of plasmonic assemblies in the fields of biosensing, nonlinear optics, and photonics. Particularly, in the field of biosensing, rapid advances have occurred in the use of plasmonic assemblies for real-time single-molecule sensing. Compared to individual particles, the use of assemblies as sensors provides stronger signals, more control over the optical properties, and access to a broader range of timescales. In the past years, they have been used to directly reveal single-molecule interactions, mechanical properties, and conformational dynamics. This review summarizes the development of real-time single-molecule sensors built around plasmonic assemblies. First, a brief overview of their optical properties is given, and then recent applications are described. The current challenges in the field and suggestions to overcome those challenges are discussed in detail. Their stability, specificity, and sensitivity as sensors provide a complementary approach to other single-molecule techniques like force spectroscopy and single-molecule fluorescence. In future applications, the impact in real-time sensing on ultralong timescales (hours) and ultrashort timescales (sub-millisecond), time windows that are difficult to access using other techniques, is particularly foreseen.

Advancing Biosensors with Machine Learning

Chemometrics play a critical role in biosensors-based detection, analysis, and diagnosis. Nowadays, as a branch of artificial intelligence (AI), machine learning (ML) have achieved impressive advances. However, novel advanced ML methods, especially deep learning, which is famous for image analysis, facial recognition, and speech recognition, has remained relatively elusive to the biosensor community. Herein, how ML can be beneficial to biosensors is systematically discussed. The advantages and drawbacks of most popular ML algorithms are summarized on the basis of sensing data analysis. Specially, deep learning methods such as convolutional neural network (CNN) and recurrent neural network (RNN) are emphasized. Diverse ML-assisted electrochemical biosensors, wearable electronics, SERS and other spectra-based biosensors, fluorescence biosensors and colorimetric biosensors are comprehensively discussed. Furthermore, biosensor networks and multibiosensor data fusion are introduced. This review will nicely bridge ML with biosensors, and greatly expand chemometrics for detection, analysis, and diagnosis.

Machine learning-enabled multiplexed microfluidic sensors

High-throughput, cost-effective, and portable devices can enhance the performance of point-of-care tests. Such devices are able to acquire images from samples at a high rate in combination with microfluidic chips in point-of-care applications. However, interpreting and analyzing the large amount of acquired data is not only a labor-intensive and time-consuming process, but also prone to the bias of the user and low accuracy. Integrating machine learning (ML) with the image acquisition capability of smartphones as well as increasing computing power could address the need for high-throughput, accurate, and automatized detection, data processing, and quantification of results. Here, ML-supported diagnostic technologies are presented. These technologies include quantification of colorimetric tests, classification of biological samples (cells and sperms), soft sensors, assay type detection, and recognition of the fluid properties. Challenges regarding the implementation of ML methods, including the required number of data points, image acquisition prerequisites, and execution of data-limited experiments are also discussed.

Unravelling the Kinetic Model of Photochemical Reactions via Deep Learning

Time-resolved spectroscopies have been playing an essential role in the elucidation of the fundamental mechanisms of light-driven processes, particularly in exploring relaxation models for electronically excited molecules. However, the determination of such models from experimentally obtained time-resolved and spectrally resolved data still demands a high degree of intuition, frequently poses numerical challenges, and is often not free from ambiguities. Here, we demonstrate the analysis of time-resolved laser spectroscopy data via a deep learning network to obtain the correct relaxation kinetic model. In its current design, the presented Deep Spectroscopy Kinetic Analysis Network (DeepSKAN) can predict kinetic models (involved states and relaxation pathways) consisting of up to five states, which results in 103 possible different classes, by estimating the probability of occurrence of a given kinetic model class. DeepSKAN was trained with synthetic time-resolved spectra spanning over 4 orders of magnitude in time with a unitless time axis, thereby demonstrating its potential as a universal approach for analyzing data from various time-resolved spectroscopy techniques in different time ranges. By adding the probabilities of each pathway of the top-k models normalized by the total probability, we can determine the relaxation pathways for a given data set with high certainty (up to 99%). Due to its architecture and training, DeepSKAN is robust against experimental noise and typical preanalysis errors like time-zero corrections. Application of DeepSKAN to experimental data is successfully demonstrated for three different photoinduced processes: transient absorption of the retinal isomerization, transient IR spectroscopy of the relaxation of the photoactivated DRONPA, and transient absorption of the dynamics in lycopene. This approach delivers kinetic models and could be a unifying asset in several areas of spectroscopy.

Using Deep Learning to Identify Molecular Junction Characteristics

The scanning tunneling microscope-based break junction (STM-BJ) is used widely to create and characterize single metal-molecule-metal junctions. In this technique, conductance is continuously recorded as a metal point contact is broken in a solution of molecules. Conductance plateaus are seen when stable molecular junctions are formed. Typically, thousands of junctions are created and measured, yielding thousands of distinct conductance versus extension traces. However, such traces are rarely analyzed individually to recognize the types of junctions formed. Here, we present a deep learning-based method to identify molecular junctions and show that it performs better than several commonly used and recently reported techniques. We demonstrate molecular junction identification from mixed solution measurements with accuracies as high as 97%. We also apply this model to an in situ electric field-driven isomerization reaction of a [3]cumulene to follow the reaction over time. Furthermore, we demonstrate that our model can remain accurate even when a key parameter, the average junction conductance, is eliminated from the analysis, showing that our model goes beyond conventional analysis in existing methods.

Deep-Channel uses deep neural networks to detect single-molecule events from patch-clamp data

Single-molecule research techniques such as patch-clamp electrophysiology deliver unique biological insight by capturing the movement of individual proteins in real time, unobscured by whole-cell ensemble averaging. The critical first step in analysis is event detection, so called "idealisation", where noisy raw data are turned into discrete records of protein movement. To date there have been practical limitations in patch-clamp data idealisation; high quality idealisation is typically laborious and becomes infeasible and subjective with complex biological data containing many distinct native single-ion channel proteins gating simultaneously. Here, we show a deep learning model based on convolutional neural networks and long short-term memory architecture can automatically idealise complex single molecule activity more accurately and faster than traditional methods. There are no parameters to set; baseline, channel amplitude or numbers of channels for example. We believe this approach could revolutionise the unsupervised automatic detection of single-molecule transition events in the future.

Study of the Application of Deep Convolutional Neural Networks (CNNs) in Processing Sensor Data and Biomedical Images

The fast progress in research and development of multifunctional, distributed sensor networks has brought challenges in processing data from a large number of sensors. Using deep learning methods such as convolutional neural networks (CNN), it is possible to build smarter systems to forecasting future situations as well as precisely classify large amounts of data from sensors. Multi-sensor data from atmospheric pollutants measurements that involves five criteria, with the underlying analytic model unknown, need to be categorized, so do the Diabetic Retinopathy (DR) fundus images dataset. In this work, we created automatic classifiers based on a deep convolutional neural network (CNN) with two models, a simpler feedforward model with dual modules and an Inception Resnet v2 model, and various structural tweaks for classifying the data from the two tasks. For segregating multi-sensor data, we trained a deep CNN-based classifier on an image dataset extracted from the data by a novel image generating method. We created two deepened and one reductive feedforward network for DR phase classification. The validation accuracies and visualization results show that increasing deep CNN structure depth or kernels number in convolutional layers will not indefinitely improve the classification quality and that a more sophisticated model does not necessarily achieve higher performance when training datasets are quantitatively limited, while increasing training image resolution can induce higher classification accuracies for trained CNNs. The methodology aims at providing support for devising classification networks powering intelligent sensors.

Single Molecule Identification and Quantification of Glycosaminoglycans Using Solid-State Nanopores

Glycosaminoglycans (GAGs) are a class of polysaccharides with potent biological activities. Due to their complex and heterogeneous composition, varied charge, polydispersity, and presence of isobaric stereoisomers, the analysis of GAG samples poses considerable challenges to current analytical techniques. In the present study, we combined solid-state nanopores-a single molecule sensor with a support vector machine (SVM)-a machine learning algorithm for the analysis of GAGs. Our results indicate that the nanopore/SVM technique could distinguish between monodisperse fragments of heparin and chondroitin sulfate with high accuracy (>90%), allowing as low as 0.8% (w/w) of chondroitin sulfate impurities in a heparin sample to be detected. In addition, the nanopore/SVM technique distinguished between unfractionated heparin (UFH) and enoxaparin (low molecular weight heparin) with an accuracy of ∼94% on average. With a reference sample for calibration, a nanopore could achieve nanomolar sensitivity and a 5-Log dynamic range. We were able to quantify heparin with reasonable accuracy using multiple nanopores. Our studies demonstrate the potential of the nanopore/SVM technique to quantify and identify GAGs.

Mimicking synaptic functionality with an InAs nanowire phototransistor

We demonstrate a nanowire (NW) phototransistor with synaptic behavior based on inherent persistent photoconductivity. The device is comprised of a single crystalline InAs NW, covered by a native indium oxide layer acting as the photogating layer (PGL). In the negative photoresponse range, the device mimics synaptic neuromorphic behaviors of short-term plasticity, long-term plasticity (LTP), and paired-pulse facilitation. Moreover, the transition from short-term to LTP is observed as the stimulus intensity increases, behaving in accord with the feature of cooperativity. The synaptic behaviors of the device are attributed to the photo-generated electrons trapped/detrapped in the PGL. This NW-based photonic synaptic device would find promising applications in neuromorphic systems and networks.

A New Frontier: The Convergence of Nanotechnology, Brain Machine Interfaces, and Artificial Intelligence

A confluence of technological capabilities is creating an opportunity for machine learning and artificial intelligence (AI) to enable "smart" nanoengineered brain machine interfaces (BMI). This new generation of technologies will be able to communicate with the brain in ways that support contextual learning and adaptation to changing functional requirements. This applies to both invasive technologies aimed at restoring neurological function, as in the case of neural prosthesis, as well as non-invasive technologies enabled by signals such as electroencephalograph (EEG). Advances in computation, hardware, and algorithms that learn and adapt in a contextually dependent way will be able to leverage the capabilities that nanoengineering offers the design and functionality of BMI. We explore the enabling capabilities that these devices may exhibit, why they matter, and the state of the technologies necessary to build them. We also discuss a number of open technical challenges and problems that will need to be solved in order to achieve this.

Recognition Tunneling of Canonical and Modified RNA Nucleotides for Their Identification with the Aid of Machine Learning

In the present study, we demonstrate a tunneling nanogap technique to identify individual RNA nucleotides, which can be used as a mechanism to read the nucleobases for direct sequencing of RNA in a solid-state nanopore. The tunneling nanogap is composed of two electrodes separated by a distance of <3 nm and functionalized with a recognition molecule. When a chemical entity is captured in the gap, it generates electron tunneling currents, a process we call recognition tunneling (RT). Using RT nanogaps created in a scanning tunneling microscope (STM), we acquired the electron tunneling signals for the canonical and two modified RNA nucleotides. To call the individual RNA nucleotides from the RT data, we adopted a machine learning algorithm, support vector machine (SVM), for the data analysis. Through the SVM, we were able to identify the individual RNA nucleotides and distinguish them from their DNA counterparts with reasonably high accuracy. Since each RNA nucleoside contains a hydroxyl group at the 2'-position of its sugar ring in an RNA strand, it allows for the formation of a tunneling junction at a larger nanogap compared to the DNA nucleoside in a DNA strand, which lacks the 2' hydroxyl group. It also proves advantageous for the manufacture of RT devices. This study is a proof-of-principle demonstration for the development of an RT nanopore device for directly sequencing single RNA molecules, including those bearing modifications.

QuipuNet: Convolutional Neural Network for Single-Molecule Nanopore Sensing

Nanopore sensing is a versatile technique for the analysis of molecules on the single-molecule level. However, extracting information from data with established algorithms usually requires time-consuming checks by an experienced researcher due to inherent variability of solid-state nanopores. Here, we develop a convolutional neural network (CNN) for the fully automated extraction of information from the time-series signals obtained by nanopore sensors. In our demonstration, we use a previously published data set on multiplexed single-molecule protein sensing. The neural network learns to classify translocation events with greater accuracy than previously possible, while also increasing the number of analyzable events by a factor of 5. Our results demonstrate that deep learning can achieve significant improvements in single molecule nanopore detection with potential applications in rapid diagnostics.

Automating drug discovery

Small-molecule drug discovery can be viewed as a challenging multidimensional problem in which various characteristics of compounds - including efficacy, pharmacokinetics and safety - need to be optimized in parallel to provide drug candidates. Recent advances in areas such as microfluidics-assisted chemical synthesis and biological testing, as well as artificial intelligence systems that improve a design hypothesis through feedback analysis, are now providing a basis for the introduction of greater automation into aspects of this process. This could potentially accelerate time frames for compound discovery and optimization and enable more effective searches of chemical space. However, such approaches also raise considerable conceptual, technical and organizational challenges, as well as scepticism about the current hype around them. This article aims to identify the approaches and technologies that could be implemented robustly by medicinal chemists in the near future and to critically analyse the opportunities and challenges for their more widespread application.