Top-down machine learning approach for high-throughput single-molecule analysis

FluoroTensor: Identification and tracking of colocalised molecules and their stoichiometries in multi-colour single molecule imaging via deep learning

The identification of photobleaching steps in single molecule fluorescence imaging is a well-established procedure for analysing the stoichiometries of molecular complexes. Nonetheless, the method is challenging with protein fluorophores because of the high levels of noise, rapid bleaching and highly variable signal intensities, all of which complicate methods based on statistical analyses of intensities to identify bleaching steps. It has recently been shown that deep learning by convolutional neural networks can yield an accurate analysis with a relatively short computational time. We describe here an improved use of such an approach that detects bleaching events even in the first time point of observation, and we have included this within an integrated software package incorporating fluorescence spot detection, colocalisation, tracking, FRET and photobleaching step analyses of single molecules or complexes. This package, known as FluoroTensor, is written in Python with a self-explanatory user interface.

Efficient, nonparametric removal of noise and recovery of probability distributions from time series using nonlinear-correlation functions: Photon and photon-counting noise

A preceding paper [M. Dhar, J. A. Dickinson, and M. A. Berg, J. Chem. Phys. 159, 054110 (2023)] shows how to remove additive noise from an experimental time series, allowing both the equilibrium distribution of the system and its Green's function to be recovered. The approach is based on nonlinear-correlation functions and is fully nonparametric: no initial model of the system or of the noise is needed. However, single-molecule spectroscopy often produces time series with either photon or photon-counting noise. Unlike additive noise, photon noise is signal-size correlated and quantized. Photon counting adds the potential for bias. This paper extends noise-corrected-correlation methods to these cases and tests them on synthetic datasets. Neither signal-size correlation nor quantization is a significant complication. Analysis of the sampling error yields guidelines for the data quality needed to recover the properties of a system with a given complexity. We show that bias in photon-counting data can be corrected, even at the high count rates needed to optimize the time resolution. Using all these results, we discuss the factors that limit the time resolution of single-molecule spectroscopy and the conditions that would be needed to push measurements into the submicrosecond region.

A Sequential Binding Mechanism for 5' Splice Site Recognition and Modulation for the Human U1 snRNP

Splice site recognition is essential for defining the transcriptome. Drugs like risdiplam and branaplam change how U1 snRNP recognizes particular 5' splice sites (5'SS) and promote U1 snRNP binding and splicing at these locations. Despite the therapeutic potential of 5'SS modulators, the complexity of their interactions and snRNP substrates have precluded defining a mechanism for 5'SS modulation. We have determined a sequential binding mechanism for modulation of -1A bulged 5'SS by branaplam using a combination of ensemble kinetic measurements and colocalization single molecule spectroscopy (CoSMoS). Our mechanism establishes that U1-C protein binds reversibly to U1 snRNP, and branaplam binds to the U1 snRNP/U1-C complex only after it has engaged a -1A bulged 5'SS. Obligate orders of binding and unbinding explain how reversible branaplam interactions cause formation of long-lived U1 snRNP/5'SS complexes. Branaplam is a ribonucleoprotein, not RNA duplex alone, targeting drug whose action depends on fundamental properties of 5'SS recognition.

cAMP binding to closed pacemaker ion channels is cooperative

The cooperative action of the subunits in oligomeric receptors enables fine-tuning of receptor activation, as demonstrated for the regulation of voltage-activated HCN pacemaker ion channels by relating cAMP binding to channel activation in ensemble signals. HCN channels generate electric rhythmicity in specialized brain neurons and cardiomyocytes. There is conflicting evidence on whether binding cooperativity does exist independent of channel activation or not, as recently reported for detergent-solubilized receptors positioned in zero-mode waveguides. Here, we show positive cooperativity in ligand binding to closed HCN2 channels in native cell membranes by following the binding of individual fluorescence-labeled cAMP molecules. Kinetic modeling reveals that the affinity of the still empty binding sites rises with increased degree of occupation and that the transition of the channel to a flip state is promoted accordingly. We conclude that ligand binding to the subunits in closed HCN2 channels not pre-activated by voltage is already cooperative. Hence, cooperativity is not causally linked to channel activation by voltage. Our analysis also shows that single-molecule binding measurements at equilibrium can quantify cooperativity in ligand binding to receptors in native membranes.

Strategies for Overcoming the Single-Molecule Concentration Barrier

Fluorescence-based single-molecule approaches have helped revolutionize our understanding of chemical and biological mechanisms. Unfortunately, these methods are only suitable at low concentrations of fluorescent molecules so that single fluorescent species of interest can be successfully resolved beyond background signal. The application of these techniques has therefore been limited to high-affinity interactions despite most biological and chemical processes occurring at much higher reactant concentrations. Fortunately, recent methodological advances have demonstrated that this concentration barrier can indeed be broken, with techniques reaching concentrations as high as 1 mM. The goal of this Review is to discuss the challenges in performing single-molecule fluorescence techniques at high-concentration, offer applications in both biology and chemistry, and highlight the major milestones that shatter the concentration barrier. We also hope to inspire the widespread use of these techniques so we can begin exploring the new physical phenomena lying beyond this barrier.

Methods for automating the analysis of live-cell single-molecule FRET data

Single-molecule FRET (smFRET) is a powerful imaging platform capable of revealing dynamic changes in the conformation and proximity of biological molecules. The expansion of smFRET imaging into living cells creates both numerous new research opportunities and new challenges. Automating dataset curation processes is critical to providing consistent, repeatable analysis in an efficient manner, freeing experimentalists to advance the technical boundaries and throughput of what is possible in imaging living cells. Here, we devise an automated solution to the problem of multiple particles entering a region of interest, an otherwise labor-intensive and subjective process that had been performed manually in our previous work. The resolution of these two issues increases the quantity of FRET data and improves the accuracy with which FRET distributions are generated, increasing knowledge about the biological functions of the molecules under study. Our automated approach is straightforward, interpretable, and requires only localization and intensity values for donor and acceptor channel signals, which we compute through our previously published smCellFRET pipeline. The development of our automated approach is informed by the insights of expert experimentalists with extensive experience inspecting smFRET trajectories (displacement and intensity traces) from live cells. We test our automated approach against our recently published research on the metabotropic glutamate receptor 2 (mGluR2) and reveal substantial similarities, as well as potential shortcomings in the manual curation process that are addressable using the algorithms we developed here.

Mechanism of ADP-Inhibited ATP Hydrolysis in Single Proton-Pumping FoF1-ATP Synthase Trapped in Solution

F_oF₁-ATP synthases in mitochondria, in chloroplasts, and in most bacteria are proton-driven membrane enzymes that supply the cells with ATP made from ADP and phosphate. Different control mechanisms exist to monitor and prevent the enzymes' reverse chemical reaction of fast wasteful ATP hydrolysis, including mechanical or redox-based blockade of catalysis and ADP inhibition. In general, product inhibition is expected to slow down the mean catalytic turnover. Biochemical assays are ensemble measurements and cannot discriminate between a mechanism affecting all enzymes equally or individually. For example, all enzymes could work more slowly at a decreasing substrate/product ratio, or an increasing number of individual enzymes could be completely blocked. Here, we examined the effect of increasing amounts of ADP on ATP hydrolysis of single Escherichia coli F_oF₁-ATP synthases in liposomes. We observed the individual catalytic turnover of the enzymes one after another by monitoring the internal subunit rotation using single-molecule Förster resonance energy transfer (smFRET). Observation times of single FRET-labeled F_oF₁-ATP synthases in solution were extended up to several seconds using a confocal anti-Brownian electrokinetic trap (ABEL trap). By counting active versus inhibited enzymes, we revealed that ADP inhibition did not decrease the catalytic turnover of all F_oF₁-ATP synthases equally. Instead, increasing ADP in the ADP/ATP mixture reduced the number of remaining active enzymes that operated at similar catalytic rates for varying substrate/product ratios.

Toward high-throughput oligomer detection and classification for early-stage aggregation of amyloidogenic protein

Aggregation kinetics of proteins and peptides have been studied extensively due to their significance in many human diseases, including neurodegenerative disorders, and the roles they play in some key physiological processes. However, most of these studies have been performed as bulk measurements using Thioflavin T or other fluorescence turn-on reagents as indicators of fibrillization. Such techniques are highly successful in making inferences about the nucleation and growth mechanism of fibrils, yet cannot directly measure assembly reactions at low protein concentrations which is the case for amyloid-β (Aβ) peptide under physiological conditions. In particular, the evolution from monomer to low-order oligomer in early stages of aggregation cannot be detected. Single-molecule methods allow direct access to such fundamental information. We developed a high-throughput protocol for single-molecule photobleaching experiments using an automated fluorescence microscope. Stepwise photobleaching analysis of the time profiles of individual foci allowed us to determine stoichiometry of protein oligomers and probe protein aggregation kinetics. Furthermore, we investigated the potential application of supervised machine learning with support vector machines (SVMs) as well as multilayer perceptron (MLP) artificial neural networks to classify bleaching traces into stoichiometric categories based on an ensemble of measurable quantities derivable from individual traces. Both SVM and MLP models achieved a comparable accuracy of more than 80% against simulated traces up to 19-mer, although MLP offered considerable speed advantages, thus making it suitable for application to high-throughput experimental data. We used our high-throughput method to study the aggregation of Aβ₄₀ in the presence of metal ions and the aggregation of α-synuclein in the presence of gold nanoparticles.

Multi-step recognition of potential 5' splice sites by the Saccharomyces cerevisiae U1 snRNP

In eukaryotes, splice sites define the introns of pre-mRNAs and must be recognized and excised with nucleotide precision by the spliceosome to make the correct mRNA product. In one of the earliest steps of spliceosome assembly, the U1 small nuclear ribonucleoprotein (snRNP) recognizes the 5' splice site (5' SS) through a combination of base pairing, protein-RNA contacts, and interactions with other splicing factors. Previous studies investigating the mechanisms of 5' SS recognition have largely been done in vivo or in cellular extracts where the U1/5' SS interaction is difficult to deconvolute from the effects of trans-acting factors or RNA structure. In this work we used colocalization single-molecule spectroscopy (CoSMoS) to elucidate the pathway of 5' SS selection by purified yeast U1 snRNP. We determined that U1 reversibly selects 5' SS in a sequence-dependent, two-step mechanism. A kinetic selection scheme enforces pairing at particular positions rather than overall duplex stability to achieve long-lived U1 binding. Our results provide a kinetic basis for how U1 may rapidly surveil nascent transcripts for 5' SS and preferentially accumulate at these sequences rather than on close cognates.

Deciphering a hexameric protein complex with Angstrom optical resolution

Cryogenic optical localization in three dimensions (COLD) was recently shown to resolve up to four binding sites on a single protein. However, because COLD relies on intensity fluctuations that result from the blinking behavior of fluorophores, it is limited to cases where individual emitters show different brightness. This significantly lowers the measurement yield. To extend the number of resolved sites as well as the measurement yield, we employ partial labeling and combine it with polarization encoding in order to identify single fluorophores during their stochastic blinking. We then use a particle classification scheme to identify and resolve heterogenous subsets and combine them to reconstruct the three-dimensional arrangement of large molecular complexes. We showcase this method (polarCOLD) by resolving the trimer arrangement of proliferating cell nuclear antigen (PCNA) and six different sites of the hexamer protein Caseinolytic Peptidase B (ClpB) of Thermus thermophilus in its quaternary structure, both with Angstrom resolution. The combination of polarCOLD and single-particle cryogenic electron microscopy (cryoEM) promises to provide crucial insight into intrinsic heterogeneities of biomolecular structures. Furthermore, our approach is fully compatible with fluorescent protein labeling and can, thus, be used in a wide range of studies in cell and membrane biology.

Gold Ion Beam Milled Gold Zero-Mode Waveguides

Zero-mode waveguides (ZMWs) are widely used in single molecule fluorescence microscopy for their enhancement of emitted light and the ability to study samples at physiological concentrations. ZMWs are typically produced using photo or electron beam lithography. We report a new method of ZMW production using focused ion beam (FIB) milling with gold ions. We demonstrate that ion-milled gold ZMWs with 200 nm apertures exhibit similar plasmon-enhanced fluorescence seen with ZMWs fabricated with traditional techniques such as electron beam lithography.

Exploring molecular biology in sequence space: The road to next-generation single-molecule biophysics

Next-generation sequencing techniques have led to a new quantitative dimension in the biological sciences. In particular, integrating sequencing techniques with biophysical tools allows sequence-dependent mechanistic studies. Using the millions of DNA clusters that are generated during sequencing to perform high-throughput binding affinity and kinetics measurements enabled the construction of energy landscapes in sequence space, uncovering relationships between sequence, structure, and function. Here, we review the approaches to perform ensemble fluorescence experiments on next-generation sequencing chips for variations of DNA, RNA, and protein sequences. As the next step, we anticipate that these fluorescence experiments will be pushed to the single-molecule level, which can directly uncover kinetics and molecular heterogeneity in an unprecedented high-throughput fashion. Molecular biophysics in sequence space, both at the ensemble and single-molecule level, leads to new mechanistic insights. The wide spectrum of applications in biology and medicine ranges from the fundamental understanding of evolutionary pathways to the development of new therapeutics.

Biomolecular Binding under Confinement: Statistical Predictions of Steric Influence in Absence of Long-Distance Interactions

We propose a theoretical model for the influence of confinement on biomolecular binding at the single-molecule scale at equilibrium, based on the change of the number of microstates (localization and orientation) upon reaction. Three cases are discussed: DNA sequences shorter and longer than the single strain DNA Kuhn length and spherical proteins, confined into a spherical container (liposome, droplet, etc.). The influence of confinement is found to be highly dependent on the molecular structure and significant for large molecules (relative to container size).

A viral genome packaging ring-ATPase is a flexibly coordinated pentamer

Multi-subunit ring-ATPases carry out a myriad of biological functions, including genome packaging in viruses. Though the basic structures and functions of these motors have been well-established, the mechanisms of ATPase firing and motor coordination are poorly understood. Here, using single-molecule fluorescence, we determine that the active bacteriophage T4 DNA packaging motor consists of five subunits of gp17. By systematically doping motors with an ATPase-defective subunit and selecting single motors containing a precise number of active or inactive subunits, we find that the packaging motor can tolerate an inactive subunit. However, motors containing one or more inactive subunits exhibit fewer DNA engagements, a higher failure rate in encapsidation, reduced packaging velocity, and increased pausing. These findings suggest a DNA packaging model in which the motor, by re-adjusting its grip on DNA, can skip an inactive subunit and resume DNA translocation, suggesting that strict coordination amongst motor subunits of packaging motors is not crucial for function.

Single-molecule imaging with cell-derived nanovesicles reveals early binding dynamics at a cyclic nucleotide-gated ion channel

Ligand binding to membrane proteins is critical for many biological signaling processes. However, individual binding events are rarely directly observed, and their asynchronous dynamics are occluded in ensemble-averaged measures. For membrane proteins, single-molecule approaches that resolve these dynamics are challenged by dysfunction in non-native lipid environments, lack of access to intracellular sites, and costly sample preparation. Here, we introduce an approach combining cell-derived nanovesicles, microfluidics, and single-molecule fluorescence colocalization microscopy to track individual binding events at a cyclic nucleotide-gated TAX-4 ion channel critical for sensory transduction. Our observations reveal dynamics of both nucleotide binding and a subsequent conformational change likely preceding pore opening. Kinetic modeling suggests that binding of the second ligand is either independent of the first ligand or exhibits up to ~10-fold positive binding cooperativity. This approach is broadly applicable to studies of binding dynamics for proteins with extracellular or intracellular domains in native cell membrane.

Unsupervised selection of optimal single-molecule time series idealization criterion

Single-molecule (SM) approaches have provided valuable mechanistic information on many biophysical systems. As technological advances lead to ever-larger data sets, tools for rapid analysis and identification of molecules exhibiting the behavior of interest are increasingly important. In many cases the underlying mechanism is unknown, making unsupervised techniques desirable. The divisive segmentation and clustering (DISC) algorithm is one such unsupervised method that idealizes noisy SM time series much faster than computationally intensive approaches without sacrificing accuracy. However, DISC relies on a user-selected objective criterion (OC) to guide its estimation of the ideal time series. Here, we explore how different OCs affect DISC’s performance for data typical of SM fluorescence imaging experiments. We find that OCs differing in their penalty for model complexity each optimize DISC’s performance for time series with different properties such as signal/noise and number of sample points. Using a machine learning approach, we generate a decision boundary that allows unsupervised selection of OCs based on the input time series to maximize performance for different types of data. This is particularly relevant for SM fluorescence data sets, which often have signal/noise near the derived decision boundary and include time series of nonuniform length because of stochastic bleaching. Our approach, AutoDISC, allows unsupervised per-molecule optimization of DISC, which will substantially assist in the rapid analysis of high-throughput SM data sets with noisy samples and nonuniform time windows.

KERA: analysis tool for multi-process, multi-state single-molecule data

Molecular machines within cells dynamically assemble, disassemble and reorganize. Molecular interactions between their components can be observed at the single-molecule level and quantified using colocalization single-molecule spectroscopy, in which individual labeled molecules are seen transiently associating with a surface-tethered partner, or other total internal reflection fluorescence microscopy approaches in which the interactions elicit changes in fluorescence in the labeled surface-tethered partner. When multiple interacting partners can form ternary, quaternary and higher order complexes, the types of spatial and temporal organization of these complexes can be deduced from the order of appearance and reorganization of the components. Time evolution of complex architectures can be followed by changes in the fluorescence behavior in multiple channels. Here, we describe the kinetic event resolving algorithm (KERA), a software tool for organizing and sorting the discretized fluorescent trajectories from a range of single-molecule experiments. KERA organizes the data in groups by transition patterns, and displays exhaustive dwell time data for each interaction sequence. Enumerating and quantifying sequences of molecular interactions provides important information regarding the underlying mechanism of the assembly, dynamics and architecture of the macromolecular complexes. We demonstrate KERA's utility by analyzing conformational dynamics of two DNA binding proteins: replication protein A and xeroderma pigmentosum complementation group D helicase.

cAMP binding to closed pacemaker ion channels is non-cooperative

Electrical activity in the brain and heart depends on rhythmic generation of action potentials by pacemaker ion channels (HCN) whose activity is regulated by cAMP binding1. Previous work has uncovered evidence for both positive and negative cooperativity in cAMP binding2,3, but such bulk measurements suffer from limited parameter resolution. Efforts to eliminate this ambiguity using single-molecule techniques have been hampered by the inability to directly monitor binding of individual ligand molecules to membrane receptors at physiological concentrations. Here we overcome these challenges using nanophotonic zero-mode waveguides4 to directly resolve binding dynamics of individual ligands to multimeric HCN1 and HCN2 ion channels. We show that cAMP binds independently to all four subunits when the pore is closed, despite a subsequent conformational isomerization to a flip state at each site. The different dynamics in binding and isomerization are likely to underlie physiologically distinct responses of each isoform to cAMP5 and provide direct validation of the ligand-induced flip-state model6-9. This approach for observing stepwise binding in multimeric proteins at physiologically relevant concentrations can directly probe binding allostery at single-molecule resolution in other intact membrane proteins and receptors.

A Python Toolbox for Unbiased Statistical Analysis of Fluorescence Intermittency of Multilevel Emitters

We report on a Python toolbox for unbiased statistical analysis of fluorescence intermittency properties of single emitters. Intermittency, that is, step-wise temporal variations in the instantaneous emission intensity and fluorescence decay rate properties, is common to organic fluorophores, II-VI quantum dots, and perovskite quantum dots alike. Unbiased statistical analysis of intermittency switching time distributions, involved levels, and lifetimes are important to avoid interpretation artifacts. This work provides an implementation of Bayesian changepoint analysis and level clustering applicable to time-tagged single-photon detection data of single emitters that can be applied to real experimental data and as a tool to verify the ramifications of hypothesized mechanistic intermittency models. We provide a detailed Monte Carlo analysis to illustrate these statistics tools and to benchmark the extent to which conclusions can be drawn on the photophysics of highly complex systems, such as perovskite quantum dots that switch between a plethora of states instead of just two.

AutoStepfinder: A fast and automated step detection method for single-molecule analysis

Single-molecule techniques allow the visualization of the molecular dynamics of nucleic acids and proteins with high spatiotemporal resolution. Valuable kinetic information of biomolecules can be obtained when the discrete states within single-molecule time trajectories are determined. Here, we present a fast, automated, and bias-free step detection method, AutoStepfinder, that determines steps in large datasets without requiring prior knowledge on the noise contributions and location of steps. The analysis is based on a series of partition events that minimize the difference between the data and the fit. A dual-pass strategy determines the optimal fit and allows AutoStepfinder to detect steps of a wide variety of sizes. We demonstrate step detection for a broad variety of experimental traces. The user-friendly interface and the automated detection of AutoStepfinder provides a robust analysis procedure that enables anyone without programming knowledge to generate step fits and informative plots in less than an hour.

•

Fast, automated, and bias-free detection of steps within single-molecule trajectories

•

Robust step detection without any prior knowledge on the data

•

A dual-pass strategy for the detection of steps over a wide variety of scales

•

A user-friendly interface for a simplified step fitting procedure

Single-molecule techniques have made it possible to track individual protein complexes in real time with a nanometer spatial resolution and a millisecond timescale. Accurate determination of the dynamic states within single-molecule time traces provides valuable kinetic information that underlie the function of biological macromolecules. Here, we present a new automated step detection method called AutoStepfinder, a versatile, robust, and easy-to-use algorithm that allows researchers to determine the kinetic states within single-molecule time trajectories without any bias.

Studying ion channel conformation dynamics by encoding coumarin as unnatural amino acid

Monitoring the conformational changes of proteins is critical to understand their function. Ion channels are membrane-bound minute machines controlling the passage of ions across biological membranes. The precise labeling of ion channels with fluorescent probes allows studying their dynamics and facilitates their characterization by high-resolution optical techniques. Here we describe a protocol for the use of a small fluorescent reporter, incorporated by expansion of the genetic code in the host cell. An important advantage of using small probes is that they are less likely to perturb protein structure, function, and trafficking. In our hands, Tyr-coumarin proved to be useful to measure the conformational changes occurring in the narrow space of the permeation pathway in single capsaicin receptors. The method described here could be directly translated to the study of membrane receptors, non-electrogenic transporters, or membrane-bound enzymes.

Learning-based event locating for single-molecule force spectroscopy

Acquiring events massively from single-molecule force spectroscopy (SMFS) experiments, which is crucial for revealing important biophysical information, is usually not straightforward. A significant amount of human labor is usually required to identify events in the measured force spectrum during measuring or before performing further data analysis. This prevents the experiment from being done in a fully-automated manner or scaling with the throughput of the measuring setup. In this work, we attempt to tackle this problem with a deep learning approach. A deep neural network model is developed to infer the occurrence of the events using the data stream from the measuring setup. We demonstrated that the proposed method could achieve high accuracy with force spectrums of a variety of samples from both optical tweezers and AFMs by learning from user-given samples instead of complicated manual algorithm designing or parameter tuning. Furthermore, we found that the trained model can be used to perform event detection on datasets measured from a different optical tweezer setup, showing the potential of being leveraged in more complex deep learning schemes.

DeepFRET, a software for rapid and automated single-molecule FRET data classification using deep learning

Single-molecule Förster Resonance energy transfer (smFRET) is an adaptable method for studying the structure and dynamics of biomolecules. The development of high throughput methodologies and the growth of commercial instrumentation have outpaced the development of rapid, standardized, and automated methodologies to objectively analyze the wealth of produced data. Here we present DeepFRET, an automated, open-source standalone solution based on deep learning, where the only crucial human intervention in transiting from raw microscope images to histograms of biomolecule behavior, is a user-adjustable quality threshold. Integrating standard features of smFRET analysis, DeepFRET consequently outputs the common kinetic information metrics. Its classification accuracy on ground truth data reached >95% outperforming human operators and commonly used threshold, only requiring ~1% of the time. Its precise and rapid operation on real data demonstrates DeepFRET's capacity to objectively quantify biomolecular dynamics and the potential to contribute to benchmarking smFRET for dynamic structural biology.

An In Vitro Single-Molecule Imaging Assay for the Analysis of Cap-Dependent Translation Kinetics

Cap-dependent protein synthesis is the predominant translation pathway in eukaryotic cells. While various biochemical and genetic approaches have allowed extensive studies of cap-dependent translation and its regulation, high resolution kinetic characterization of this translation pathway is still lacking. Recently, we developed an in vitro assay to measure cap-dependent translation kinetics with single-molecule resolution. The assay is based on fluorescently labeled antibody binding to nascent epitope-tagged polypeptide. By imaging the binding and dissociation of antibodies to and from nascent peptide-ribosome-mRNA complexes, the translation progression on individual mRNAs can be tracked. Here, we present a protocol for establishing this assay, including mRNA and PEGylated slide preparations, real-time imaging of translation, and analysis of single molecule trajectories. This assay enables tracking of individual cap-dependent translation events and resolves key translation kinetics, such as initiation and elongation rates. The assay can be widely applied to distinct translation systems and should broadly benefit in vitro studies of cap-dependent translation kinetics and translational control mechanisms.

Top-down machine learning approach for high-throughput single-molecule analysis

Single-molecule approaches provide enormous insight into the dynamics of biomolecules, but adequately sampling distributions of states and events often requires extensive sampling. Although emerging experimental techniques can generate such large datasets, existing analysis tools are not suitable to process the large volume of data obtained in high-throughput paradigms. Here, we present a new analysis platform (DISC) that accelerates unsupervised analysis of single-molecule trajectories. By merging model-free statistical learning with the Viterbi algorithm, DISC idealizes single-molecule trajectories up to three orders of magnitude faster with improved accuracy compared to other commonly used algorithms. Further, we demonstrate the utility of DISC algorithm to probe cooperativity between multiple binding events in the cyclic nucleotide binding domains of HCN pacemaker channel. Given the flexible and efficient nature of DISC, we anticipate it will be a powerful tool for unsupervised processing of high-throughput data across a range of single-molecule experiments.