Rapid Classification Research Articles

Multi- and hyperspectral classification of soft-bottom intertidal vegetation using a spectral library for coastal biodiversity remote sensing

Monitoring biodiversity and how anthropogenic pressures impact this is critical, especially as anthropogenically driven climate change continues to affect all ecosystems. Intertidal areas are exposed to particularly high levels of pressures owing to increased population density in coastal areas. Traditional methods of monitoring intertidal areas do not provide datasets with full coverage in a cost-effective or timely manner, and so the use of remote sensing to monitor these areas is becoming more common. Monitoring of ecologically important monospecific habitats, such as seagrass beds, using remote sensing techniques is well documented. However, the ability for multispectral data to distinguish efficiently and accurately between classes of vegetation with similar pigment composition, such as seagrass and green algae, has proved difficult, often requiring hyperspectral data. A machine learning approach was used to differentiate between soft-bottom intertidal vegetation classes when exposed at low tide, comparing 6 different multi- and hyperspectral remote and in situ sensors. For the library of 366 spectra, collected across Northern Europe, high accuracy (>80%) was found across all spectral resolutions. While a higher spectral resolution resulted in higher accuracy, there was no discernible increase in accuracy above 10 spectral bands (95%: Sentinel-2 MSI sensor with a spatial resolution of 20 m). This work highlights the ability of multispectral sensors to discriminate intertidal vegetation types, while also showing the most important wavelengths for this discrimination (∼530 and ∼ 730 nm), giving recommendations for spectral ranges of future satellite missions. The ability for multispectral sensors to aid in accurate and rapid intertidal vegetation classification at the taxonomic resolution of classes, could be a significant contribution for future sustainable and effective ecosystem management.

Wavelet multiscale principal components and traveling waves enabled machine learning technique for protection of MT-HVDC systems

The interconnection of renewable energy sources to conventional power grids is widely recognized and promoted by the deployment of high voltage direct current (HVDC) technology. One of the biggest challenges associated with the HVDC systems is the rapid rate of rising DC fault currents, resulting in the immediate collapse of the converter stations. Conventional schemes of AC protection and low-speed DC protection are not worthy and mature because of the abrupt rising characteristics of DC currents. Therefore, protection techniques are being developed for reliable bulk power transfer and interconnection of unsynchronized grids via DC links. Timely identification, classification, and determination of location are the most fundamental characteristics of any protection scheme. This research paper proposes a machine learning-based protection method that establishes a rapid and reliable solution addressing the drawbacks of existing protection schemes. A multi-terminal high voltage direct current (MT-HVDC) test system is designed in Matlab/Simulink. Pole-to-ground and pole-to-pole faults are applied at different fault locations to assess the validity of the proposed algorithm. For this, DC voltages and currents are measured and analyzed. Power spectral analysis of traveling waves and Wavelet multiscale principal components analysis are employed for the extraction of featured data to classify and locate the DC faults. This reduced dimensional featured data is utilized for training and testing the support vector machine learning algorithm to analyze the accuracy of the proposed protection technique. Simulation results confirm the accuracy of rapid identification, classification, and location of DC faults.

Deep learning detection and classification of gravitational waves from neutron star-black hole mergers

The Laser Interferometer Gravitational-Wave Observatory (LIGO) and Virgo Interferometer Collaborations have now detected all three classes of compact binary mergers: binary black hole (BBH), binary neutron star (BNS), and neutron star-black hole (NSBH). For coalescences involving neutron stars, the simultaneous observation of gravitational and electromagnetic radiation produced by an event, has broader potential to enhance our understanding of these events, and also to probe the equation of state (EOS) of dense matter. However, electromagnetic follow-up to gravitational wave (GW) events requires rapid real-time detection and classification of GW signals, and conventional detection approaches are computationally prohibitive for the anticipated rate of detection of next-generation GW detectors. In this work, we present the first deep learning based results of classification of GW signals from NSBH mergers in real LIGO data. We show for the first time that a deep neural network can successfully distinguish all three classes of compact binary mergers and separate them from detector noise. Specifically, we train a convolutional neural network (CNN) on ∼500,000 data samples of real LIGO noise with injected BBH, BNS, and NSBH GW signals, and we show that our network has high sensitivity and accuracy. Most importantly, we successfully recover the two confirmed NSBH events to-date (GW191219 and GW200115) and the two confirmed BNS mergers to-date (GW170817 and GW190425), together with ∼90% of all BBH candidate events from the third Gravitational Wave Transient Catalog, GWTC-3. These results are an important step towards low-latency real-time GW detection, enabling multi-messenger astronomy.

Machine learning and deep learning based on the small FT-MIR dataset for fine-grained sampling site recognition of boletus tomentipes

This study proposed the necessity of identifying the sampling sites for Boletus tomentipes (B.tomentipes) in combination with cadmium content and environmental factors. Based on fourier transform mid-infrared spectroscopy (FT-MIR) preprocessing by 1st, 2nd, MSC, SNV and SG, five machine learning (ML) algorithms (NB, DT, KNN, RF, SVM) and three Gradient Boosting Machine (GBM) algorithms (XGBoost, LightGBM, CatBoost) were built. To avoid complex preprocessing, we construct BoletusResnet model, propose the concepts of 3DCOS, 3DCOS projected images, index images in addition to 2DCOS, and combine them with deep learning (DL) for classification for the first time. It shows that GBM has higher accuracy than ML and DL has better accuracy than GBM. The four DL models presented in this paper achieve fine-grained sampling sites recognition based on small samples with 100 % accuracy, and a computer application system was developed on them. Therefore, spectral image processing combined with DL is a rapid and efficient classification method which can be widely used in food identification.

GPMeta: a GPU-accelerated method for ultrarapid pathogen identification from metagenomic sequences.

Metagenomic sequencing (mNGS) is a powerful diagnostic tool to detect causative pathogens in clinical microbiological testing owing to its unbiasedness and substantially reduced costs. Rapid and accurate classification of metagenomic sequences is a critical procedure for pathogen identification in dry-lab step of mNGS test. However, clinical practices of the testing technology are hampered by the challenge of classifying sequences within a clinically relevant timeframe. Here, we present GPMeta, a novel GPU-accelerated approach to ultrarapid pathogen identification from complex mNGS data, allowing users to bypass this limitation. Using mock microbial community datasets and public real metagenomic sequencing datasets from clinical samples, we show that GPMeta has not only higher accuracy but also significantly higher speed than existing state-of-the-art tools such as Bowtie2, Bwa, Kraken2 and Centrifuge. Furthermore, GPMeta offers GPMetaC clustering algorithm, a statistical model for clustering and rescoring ambiguous alignments to improve the discrimination of highly homologous sequences from microbial genomes with average nucleotide identity >95%. GPMetaC exhibits higher precision and recall rate than others. GPMeta underlines its key role in the development of the mNGS test in infectious diseases that require rapid turnaround times. Further study will discern how to best and easily integrate GPMeta into routine clinical practices. GPMeta is freely accessible to non-commercial users at https://github.com/Bgi-LUSH/GPMeta.

A rapid and nondestructive approach for forensic identification of novel psychoactive substances using shifted‐excitation Raman difference spectroscopy and machine learning

Rapid drug detection is vital for combating drug abuse and inhibiting transmission, and the emergence of various novel psychoactive substances has placed higher demands on field testing. To establish a rapid, non-destructive and accurate analytical and classification method for testing novel psychoactive substances, SERDS was used to detect 37 NPS including fentanyl, amphetamine and synthetic cannabinoid under the experimental conditions of laser light source (785&785.5) nm. SERDS was also used in conjunction with machine learning to find the best classification prediction model. In this study,the characteristic peaks of 37 NPS analogues were extracted, and the characteristic peaks were attributed to the structures of the substances. We compared the classification effects of SVM,KNN,Ensemble classifier, Neural Network,Tree,Naive Bayes,Discriminant Analysis, and provided three Hyperparametric optimization methods. Finally, the cross validation accuracy of SVM under BOA optimization is 97.3%, which can well distinguish three families of samples. The ability of SERDS to resist strong fluorescence interference makes it a powerful tool for on-site analysis of NPS. In this study, we have given 37 unreported standard Raman peaks of NPS and 3 seized samples, which can be used as a reference. Shift-Excitation Differential Raman spectroscopy combined with machine learning can effectively provide solutions for customs, health care, on-site police, Large event security, and trace evidence inspection.

QSP: An open sequence database for quorum sensing related gene analysis with an automatic annotation pipeline

Quorum sensing (QS) has attracted great attention due to its important role in the bacterial interactions and its relevance to water management. With the development of high-throughput sequencing technology, a specific database for QS-related sequence annotation is urgently needed. Here, Hidden Markov Model (HMM) profiles for 38 types of QS-related proteins were built using a total of 4024 collected seed sequences. Based on both homolog search and keywords confirmation against the non-redundant database, we established a QS-related protein (QSP) database, that includes 809,721 protein sequences and 186,133 nucleotide sequences, downloaded available at: https://github.com/chunxiao-dcx/QSP. The entries were classified into 38 types and 315 subtypes among 91 bacterial phyla. Furthermore, an automatic annotation pipeline, named QSAP, was developed for rapid annotation, classification and abundance quantification of QSP-like sequences from sequencing data. This pipeline provided the two homolog alignment strategies offered by Diamond (Blastp) or HMMER (Hmmscan), as well as a data cleansing function for a subset or union set of the hits. The pipeline was tested using 14 metagenomic samples from various water environments, including activated sludge, deep-sea sediments, estuary water, and reservoir water. The QSAP pipeline is freely available for academic use in the code repository at: https://github.com/chunxiao-dcx/QSAP. The establishment of this database and pipeline, provides a useful tool for QS-related sequence annotation in a wide range of projects, and will increase our understanding of QS communication in aquatic environments.

Rapid and accurate classification of Aspergillus ochraceous contamination in Robusta green coffee bean through near-infrared spectral analysis using machine learning

Near-infrared (NIR) spectral-based classification of Aspergillus ochraceous contamination in the Robusta green coffee bean was investigated. Six different learning algorithms, including linear discriminant analysis (LDA), support vector machine (SVM), k-nearest neighbors (KNN), decision tree (Tree), Naive Bayes (NB), and quadratic discriminant analysis (QDA), were applied for the investigating purpose. Four classes of fungal contamination on coffee beans, non-fungal contaminated beans on day 1 and day 3 (NCB-D1 and NCB-D3) and fungal contaminated beans on day 1 and day 3 (CB-D1 and CB-D3), were set for the classification intention. Based on the 6 learning algorithms, the Tree approach was optimal, displaying a training accuracy of 97.5%. As proven by the testing dataset, the classification accuracy of the Tree was also at 97.5%. With this number, the Tree could correctly classify 100% between the contaminated and non-contaminated coffee beans. These findings exhibit the potential of the NIR spectroscopy accompanied by machine learning for the early detection of fungal contamination in green coffee beans.

Rapid Classification of Massive Images Based on Cloud Computing Platform

Rapid Classification of Massive Images Based on Cloud Computing Platform

Raw Electroencephalogram-Based Cognitive Workload Classification Using Directed and Nondirected Functional Connectivity Analysis and Deep Learning.

With the phenomenal rise in internet-of-things devices, the use of electroencephalogram (EEG) based brain-computer interfaces (BCIs) can empower individuals to control equipment with thoughts. These allow BCI to be used and pave the way for pro-active health management and the development of internet-of-medical-things architecture. However, EEG-based BCIs have low fidelity, high variance, and EEG signals are very noisy. These challenges compel researchers to design algorithms that can process big data in real-time while being robust to temporal variations and other variations in the data. Another issue in designing a passive BCI is the regular change in user's cognitive state (measured through cognitive workload). Though considerable amount of research has been conducted on this front, methods that could withstand high variability in EEG data and still reflect the neuronal dynamics of cognitive state variations are lacking and much needed in literature. In this research, we evaluate the efficacy of a combination of functional connectivity algorithms and state-of-the-art deep learning algorithms for the classification of three different levels of cognitive workload. We acquire 64-channel EEG data from 23 participants executing the n-back task at three different levels; 1-back (low-workload condition), 2-back (medium-workload condition), and 3-back (high-workload condition). We compared two different functional connectivity algorithms, namely phase transfer entropy (PTE) and mutual information (MI). PTE is a directed functional connectivity algorithm, whereas MI is non-directed. Both methods are suitable for extracting functional connectivity matrices in real-time, which could eventually be used for rapid, robust, and efficient classification. For classification, we use the recently proposed BrainNetCNN deep learning model, designed specifically to classify functional connectivity matrices. Results reveal a classification accuracy of 92.81% with MI and BrainNetCNN and a staggering 99.50% with PTE and BrainNetCNN on test data. PTE can yield a higher classification accuracy due to its robustness to linear mixing of the data and its ability to detect functional connectivity across a range of analysis lags.

Classifying metal-binding sites with neural networks.

To advance our ability to predict impacts of the protein scaffold on catalysis, robust classification schemes to define features of proteins that will influence reactivity are needed. One of these features is a protein's metal-binding ability, as metals are critical to catalytic conversion by metalloenzymes. As a step toward realizing this goal, we used convolutional neural networks (CNNs) to enable the classification of a metal cofactor binding pocket within a protein scaffold. CNNs enable images to be classified based on multiple levels of detail in the image, from edges and corners to entire objects, and can provide rapid classification. First, six CNN models were fine-tuned to classify the 20 standard amino acids to choose a performant model for amino acid classification. This model was then trained in two parallel efforts: to classify a 2D image of the environment within a given radius of the central metal binding site, either an Fe ion or a [2Fe-2S] cofactor, with the metal visible (effort 1) or the metal hidden (effort 2). We further used two sub-classifications of the [2Fe-2S] cofactor: (1) a standard [2Fe-2S] cofactor and (2) a Rieske [2Fe-2S] cofactor. The accuracy for the model correctly identifying all three defined features was >95%, despite our perception of the increased challenge of the metalloenzyme identification. This demonstrates that machine learning methodology to classify and distinguish similar metal-binding sites, even in the absence of a visible cofactor, is indeed possible and offers an additional tool for metal-binding site identification in proteins.

Rapid identification of chemical components in vitro and in vivo of Menispermi Rhizoma by integrating UPLC-Q-TOF-MS with data post-processing strategy.

Menispermi Rhizoma (MR), the dried rhizome of Menispermum dauricum DC. (Menispermaceae), has been used to treat sore throat, enteritis, dysentery, and rheumatic arthralgia. Despite extensive research on its pharmacological effects, the chemical components in vitro and in vivo have not been thoroughly studied. To establish an efficient method for rapid classification and identification of alkaloids in MR and its preparations, as well as metabolites in vivo after oral administration of MR. Rapid identification of alkaloids and absorbed components of MR was performed using ultra-performance liquid chromatography quadrupole time-of-flight mass spectrometry (UPLC-Q-TOF-MS) coupled with UNIFI software. Moreover, the characteristic fragmentations and neutral losses of different types of alkaloids in MR were summarised to realise the rapid classification of alkaloids. A total of 55 components were unambiguously or tentatively identified in MR. Among them, 37 and 31 components were found in MR capsules and tablets, respectively. Meanwhile, 109 compounds were tentatively identified in rat plasma, urine and faeces, including 55 prototypes and 54 metabolites. Hydrogenation, hydroxylation, methylation, glucuronic acid and sulphate conjugations were the dominating metabolic fates of alkaloids. The data post-processing strategy established could greatly enhance the structural identification efficiency. The results obtained might lay the foundation for further interpretation of clinical effects, mechanism of action and quality control of MR.

Data-driven selection and spectral classification of white dwarf stars

ABSTRACT The next generation of spectroscopic surveys is expected to provide spectra for hundreds of thousands of white dwarf (WD) candidates in the upcoming years. Currently, spectroscopic classification of white dwarfs is mostly done by visual inspection, requiring substantial amounts of expert attention. We propose a data-driven pipeline for fast, automatic selection, and spectroscopic classification of WD candidates, trained using spectroscopically confirmed objects with available Gaia astrometry, photometry, and Sloan Digital Sky Survey (SDSS) spectra with signal-to-noise ratios ≥9. The pipeline selects WD candidates with improved accuracy and completeness over existing algorithms, classifies their primary spectroscopic type with ${\gtrsim}90\ \hbox{per cent}$ accuracy, and spectroscopically detects main sequence companions with similar performance. We apply our pipeline to the Gaia Data Release 3 cross-matched with the SDSS Data Release 17 (DR17), identifying 424 096 high-confidence WD candidates and providing the first catalogue of automated and quantifiable classification for 36 523 WD spectra. Both the catalogue and pipeline are made available online. Such a tool will prove particularly useful for the undergoing SDSS-V survey, allowing for rapid classification of thousands of spectra at every data release.

Strong π-Metal Interaction Enables Liquid Interfacial Nanoarray-Molecule Co-assembly for Raman Sensing of Ultratrace Fentanyl Doped in Heroin, Ketamine, Morphine, and Real Urine.

Toward the challenge on reliable determination of trace fentanyl to avoid opioid overdose death in drug crisis, here we realize rapid and direct detection of trace fentanyl in real human urine without pretreatment by a portable surface enhanced Raman spectroscopy (SERS) strategy on liquid/liquid interfacial (LLI) plasmonic arrays. It was observed that fentanyl could interact with the gold nanoparticles (GNPs) surface, facilitate the LLI self-assembly, and consequently amplify the detection sensitivity with a limit of detection (LOD) as low as 1 ng/mL in aqueous solution and 50 ng/mL spiked in urine. Furthermore, we achieve multiplex blind sample recognition and classification of ultratrace fentanyl doped in other illegal drugs, which has extremely low LODs at mass concentrations of 0.02% (2 ng in 10 μg of heroin), 0.02% (2 ng in 10 μg of ketamine), and 0.1% (10 ng in 10 μg of morphine). A logic circuit of the AND gate was constructed for automatic recognition of illegal drugs with or without fentanyl doping. The data-driven analog soft independent modeling model could quickly distinguish fentanyl-doped samples from illegal drugs with 100% specificity. Molecular dynamics (MD) simulation elucidates the underlying molecular mechanism of nanoarray-molecule co-assembly through strong π-metal interactions and the differences in the SERS signal of various drug molecules. It paves a rapid identification, quantification, and classification strategy for trace fentanyl analysis, indicating broad application prospects in response to the opioid epidemic crisis.

Complete, Fully Automatic Detection and Classification of Benign and Malignant Breast Tumors Based on CT Images Using Artificial Intelligent and Image Processing.

Breast cancer is the most common type of cancer in women, and early detection is important to significantly reduce its mortality rate. This study introduces a detection and diagnosis system that automatically detects and classifies breast tumors in CT scan images. First, the contours of the chest wall are extracted from computed chest tomography images, and two-dimensional image characteristics and three-dimensional image features, together with the application of active contours without edge and geodesic active contours methods, are used to detect, locate, and circle the tumor. Then, the computer-assisted diagnostic system extracts features, quantifying and classifying benign and malignant breast tumors using a greedy algorithm and a support vector machine. The study used 174 breast tumors for experiment and training and performed cross-validation 10 times (k-fold cross-validation) to evaluate performance of the system. The accuracy, sensitivity, specificity, and positive and negative predictive values of the system were 99.43%, 98.82%, 100%, 100%, and 98.89% respectively. This system supports the rapid extraction and classification of breast tumors as either benign or malignant, helping physicians to improve clinical diagnosis.

Using Impedance Flow Cytometry for Rapid Viability Classification of Heat-Treated Bacteria.

In the future, rapid electrical characterization of cells with impedance flow cytometry promises to be a fast and accurate method for the evaluation of cell properties. In this paper, we investigate how the conductivity of the suspending medium along with the heat exposure time affects the viability classification of heat-treated E. coli. Using a theoretical model, we show that perforation of the bacteria membrane during heat exposure changes the impedance of the bacterial cell from effectively less conducting than the suspension medium to effectively more conducting. Consequently, this results in a shift in the differential argument of the complex electrical current that can be measured with impedance flow cytometry. We observe this shift experimentally through measurements on E. coli samples with varying medium conductivity and heat exposure times. We show that increased exposure time and lower medium conductivity results in improved classification between untreated and heat-treated bacteria. The best classification was achieved with a medium conductivity of 0.045 S/m after 30 min of heat exposure.

Efficient wave type fingerprinting and filtering by six-component polarization analysis

SUMMARYWe present a technique to automatically classify the wave type of seismic phases that are recorded on a single six-component recording station (measuring both three components of translational and rotational ground motion) at the Earth’s surface. We make use of the fact that each wave type leaves a unique ’fingerprint’ in the six-component motion of the sensor (i.e. the motion is unique for each wave type). This fingerprint can be extracted by performing an eigenanalysis of the data covariance matrix, similar to conventional three-component polarization analysis. To assign a wave type to the fingerprint extracted from the data, we compare it to analytically derived six-component polarization models that are valid for pure-state plane wave arrivals. For efficient classification, we make use of the supervised machine learning method of support vector machines that is trained using data-independent, analytically derived six-component polarization models. This enables the rapid classification of seismic phases in a fully automated fashion, even for large data volumes, such as encountered in land-seismic exploration or ambient noise seismology. Once the wave-type is known, additional wave parameters (velocity, directionality and ellipticity) can be directly extracted from the six-component polarization states without the need to resort to expensive optimization algorithms. We illustrate the benefits of our approach on various real and synthetic data examples for applications such as automated phase picking, aliased ground-roll suppression in land-seismic exploration and the rapid close-to real-time extraction of surface wave dispersion curves from single-station recordings of ambient noise. Additionally, we argue that an initial step of wave type classification is necessary in order to successfully apply the common technique of extracting phase velocities from combined measurements of rotational and translational motion.

Label-free, fast electronic classification of bacteria.

Microbiomes are communities of bacteria that have come into focus recently due to their major roles in human health and disease. Microbiome classification in patients will be important for the future of personalized medicine, which will require repeatable, rapid classification of microbiomes to track microbiome treatments. Current microbiome classification methods rely on bacterial genome sequencing, which has a relatively slow turnaround time and provides no physical information about the bacteria. To facilitate rapid microbiome classification, we have developed a high throughput approach to detect and classify bacteria based on their physical properties such as shape, size, and electrochemical surface characteristics. The method utilizes resistive-pulse technique and a microchannel whose width undulates on a sub-micrometer scale, establishing sensitivity not only to the object volume but also shape. As individual bacteria pass through a device, a transient decrease of the transmembrane current, called a pulse, is recorded; the current signal within each pulse contains local minima and maxima that correspond to channel undulations. Elongated bacteria spread over multiple undulations produced ‘smoothed out’ signals compared to shorter bacteria pulses. The method also allows discrimination between bacteria with the same volume but different shapes and provides quantitative information about the surface charge properties of bacteria. We have demonstrated the application of our method to perform in-situ tests of efficacy of the antibiotic cephalexin on E. coli, which causes cell elongation. The resistive-pulse measurements distinguished cephalexin-treated bacteria from untreated bacteria based on cell shape and discriminated between E. coli and S. aureus cells. These findings demonstrate that our method can rapidly discriminate between antibiotic-resistant and antibiotic-sensitive bacteria and between different bacterial species. The ability to rapidly characterize the antibiotic resistance level and bacterial species could be an extremely useful tool for biomedical studies and healthcare settings.

Optical photothermal infrared spectroscopy: A novel solution for rapid identification of antimicrobial resistance at the single-cell level via deuterium isotope labeling.

The rise and extensive spread of antimicrobial resistance (AMR) has become a growing concern, and a threat to the environment and human health globally. The majority of current AMR identification methods used in clinical setting are based on traditional microbiology culture-dependent techniques which are time-consuming or expensive to be implemented, thus appropriate antibiotic stewardship is provided retrospectively which means the first line of treatment is to hope that a broad-spectrum antibiotic works. Hence, culture-independent and single-cell technologies are needed to allow for rapid detection and identification of antimicrobial-resistant bacteria and to support a more targeted and effective antibiotic therapy preventing further development and spread of AMR. In this study, for the first time, a non-destructive phenotyping method of optical photothermal infrared (O-PTIR) spectroscopy, coupled with deuterium isotope probing (DIP) and multivariate statistical analysis was employed as a metabolic fingerprinting approach to detect AMR in Uropathogenic Escherichia coli (UPEC) at both single-cell and population levels. Principal component-discriminant function analysis (PC-DFA) of FT-IR and O-PTIR spectral data showed clear clustering patterns as a result of distinctive spectral shifts (C-D signature peaks) originating from deuterium incorporation into bacterial cells, allowing for rapid detection and classification of sensitive and resistant isolates at the single-cell level. Furthermore, the single-frequency images obtained using the C-D signature peak at 2,163 cm-1 clearly displayed the reduced ability of the trimethoprim-sensitive strain for incorporating deuterium when exposed to this antibiotic, compared to the untreated condition. Hence, the results of this study indicated that O-PTIR can be employed as an efficient tool for the rapid detection of AMR at the single-cell level.

Making the undetectable detectable: Pushing limits of biomolecular recognition by ultra-sensitive multiplexed single-particle liposome readouts.

Biomolecular recognition is fundamental for all biological and chemical processes. However, current screening methods are often reliant on large robotic devices, require long running times and much manpower as well as excessive materials cost. Here I will present our recent work implementing multiple miniaturized screening assays for ultra-sensitive and high-throughput readout at the single-particle level of DNA-DNA recognition, sub-attoliter cargo delivery, transporter function and label-free DNA biosensing. I will focus on the single-particle combinatorial multiplexed liposome fusion mediated by DNA for parallelized multistep and non-deterministic fusion of individual sub-attoliter nanocontainers (SPARCLD). Our new toolbox allowed us to attain efficient (>93%) and leakage free stochastic fusion sequences for arrays of surface-tethered target liposomes with six freely diffusing populations of cargo liposomes, each functionalized with individual lipidated single-stranded DNA and fluorescently barcoded by a distinct ratio of chromophores. Because each lipid nanoparticle is an autonomous nanocontainer the stochastic fusion sequence resulted in massive number of distinct permutations of cargo delivery loadings and sequences. Real-time recordings and machine learning analysis allowed the rapid classification of >16,000 fusions and 566 distinct fusion sequences within hours. The high-density arrays of surface-tethered target nanocontainers (∼42,000 containers per mm2) offers entire combinatorial multiplex screens using only picograms of material and is implanted for screening transporter function and label free DNA biosensing.