High-Speed Atomic Force Microscopy Imaging of DNA Three-Point-Star Motif Self Assembly Using Photothermal Off-Resonance Tapping.

High-speed atomic force microscopy (HS-AFM) is a popular molecular imaging technique for visualizing single-molecule biological processes in real-time due to its ability to image under physiological conditions in liquid environments. The photothermal off-resonance tapping (PORT) mode uses a drive laser to oscillate the cantilever in a controlled manner. This direct cantilever actuation is effective in the MHz range. Combined with operating the feedback loop on the time domain force curve rather than the resonant amplitude, PORT enables high-speed imaging at up to ten frames per second with direct control over tip-sample forces. PORT has been shown to enable imaging of delicate assembly dynamics and precise monitoring of patterns formed by biomolecules. Thus far, the technique has been used for a variety of dynamic in vitro studies, including the DNA 3-point-star motif assembly patterns shown in this work. Through a series of experiments, this protocol systematically identifies the optimal imaging parameter settings and ultimate limits of the HS-PORT AFM imaging system and how they affect biomolecular assembly processes. Additionally, it investigates potential undesired thermal effects induced by the drive laser on the sample and surrounding liquid, particularly when the scanning is limited to small areas. These findings provide valuable insights that will drive the advancement of PORT mode's application in studying complex biological systems.

Interplay of the mechanical and structural properties of DNA nanostructures determines their electrostatic interactions with lipid membranes.

Nucleic acids and lipids function in close proximity in biological processes, as well as in nanoengineered constructs for therapeutic applications. As both molecules carry a rich charge profile, and frequently coexist in complex ionic solutions, the electrostatics surely play a pivotal role in interactions between them. Here we discuss how each component of a DNA/ion/lipid system determines its electrostatic attachment. We examine membrane binding of a library of DNA molecules varying from nanoengineered DNA origami through plasmids to short DNA domains, demonstrating the interplay between the molecular structure of the nucleic acid and the phase of lipid bilayers. Furthermore, the magnitude of DNA/lipid interactions is tuned by varying the concentration of magnesium ions in the physiologically relevant range. Notably, we observe that the structural and mechanical properties of DNA are critical in determining its attachment to lipid bilayers and demonstrate that binding is correlated positively with the size, and negatively with the flexibility of the nucleic acid. The findings are utilized in a proof-of-concept comparison of membrane interactions of two DNA origami designs - potential nanotherapeutic platforms - showing how the results can have a direct impact on the choice of DNA geometry for biotechnological applications.

Conservative Policy Construction Using Variational Autoencoders for Logged Data With Missing Values.

In high-stakes applications of data-driven decision-making such as healthcare, it is of paramount importance to learn a policy that maximizes the reward while avoiding potentially dangerous actions when there is uncertainty. There are two main challenges usually associated with this problem. First, learning through online exploration is not possible due to the critical nature of such applications. Therefore, we need to resort to observational datasets with no counterfactuals. Second, such datasets are usually imperfect, additionally cursed with missing values in the attributes of features. In this article, we consider the problem of constructing personalized policies using logged data when there are missing values in the attributes of features in both training and test data. The goal is to recommend an action (treatment) when ~ X , a degraded version of X with missing values, is observed. We consider three strategies for dealing with missingness. In particular, we introduce the conservative strategy where the policy is designed to safely handle the uncertainty due to missingness. In order to implement this strategy, we need to estimate posterior distribution p(X| ~ X) and use a variational autoencoder to achieve this. In particular, our method is based on partial variational autoencoders (PVAEs) that are designed to capture the underlying structure of features with missing values.

Donor-Acceptor Stenhouse Adduct-Polydimethylsiloxane-Conjugates for Enhanced Photoswitching in Bulk Polymers.

Donor-acceptor Stenhouse adducts (DASAs) are a rapidly emerging class of visible light-activated photochromes and DASA-functionalized polymers hold great promise as biocompatible photoresponsive materials. However, the photoswitching performance of DASAs in solid polymer matrices is often low, particularly in materials below their glass transition temperature. To overcome this limitation, DASAs are conjugated to polydimethylsiloxanes which have a glass transition temperature far below room temperature and which can create a mobile molecular environment around the DASAs for achieving more solution-like photoswitching kinetics in bulk polymers. The dispersion of DASAs conjugated to such flexible oligomers into solid polymer matrices allows for more effective and tunable DASA photoswitching in stiff polymers, such as poly(methyl methacrylate), without requiring modifications of the matrix. The photoswitching of conjugates with varying polymer molecular weight, linker type, and architecture is characterized via time-dependent UV-vis spectroscopy in organic solvents and blended into polymethacrylate films. In addition, DASA-functionalized polydimethylsiloxane networks, accessible via the same synthetic route, provide an alternative solution for achieving fast and efficient DASA photoswitching in the bulk owing to their intrinsic softness and flexibility. These findings may contribute to the development of DASA-functionalized materials with better tunable, more effective, and more reversible modulation of their optical properties.

Global Bandits.

Multiarmed bandits (MABs) model sequential decision-making problems, in which a learner sequentially chooses arms with unknown reward distributions in order to maximize its cumulative reward. Most of the prior works on MAB assume that the reward distributions of each arm are independent. But in a wide variety of decision problems-from drug dosage to dynamic pricing-the expected rewards of different arms are correlated, so that selecting one arm provides information about the expected rewards of other arms as well. We propose and analyze a class of models of such decision problems, which we call global bandits (GB). In the case in which rewards of all arms are deterministic functions of a single unknown parameter, we construct a greedy policy that achieves bounded regret, with a bound that depends on the single true parameter of the problem. Hence, this policy selects suboptimal arms only finitely many times with probability one. For this case, we also obtain a bound on regret that is independent of the true parameter; this bound is sublinear, with an exponent that depends on the informativeness of the arms. We also propose a variant of the greedy policy that achieves worst case and parameter-dependent regret. Finally, we perform experiments on dynamic pricing and show that the proposed algorithms achieve significant gains with respect to the well-known benchmarks.

Functional Contour-following via Haptic Perception and Reinforcement Learning.

Many tasks involve the fine manipulation of objects despite limited visual feedback. In such scenarios, tactile and proprioceptive feedback can be leveraged for task completion. We present an approach for real-time haptic perception and decision-making for a haptics-driven, functional contour-following task: the closure of a ziplock bag. This task is challenging for robots because the bag is deformable, transparent, and visually occluded by artificial fingertip sensors that are also compliant. A deep neural net classifier was trained to estimate the state of a zipper within a robot's pinch grasp. A Contextual Multi-Armed Bandit (C-MAB) reinforcement learning algorithm was implemented to maximize cumulative rewards by balancing exploration versus exploitation of the state-action space. The C-MAB learner outperformed a benchmark Q-learner by more efficiently exploring the state-action space while learning a hard-to-code task. The learned C-MAB policy was tested with novel ziplock bag scenarios and contours (wire, rope). Importantly, this work contributes to the development of reinforcement learning approaches that account for limited resources such as hardware life and researcher time. As robots are used to perform complex, physically interactive tasks in unstructured or unmodeled environments, it becomes important to develop methods that enable efficient and effective learning with physical testbeds.

Active Learning in Context-Driven Stream Mining With an Application to Image Mining.

We propose an image stream mining method in which images arrive with contexts (metadata) and need to be processed in real time by the image mining system (IMS), which needs to make predictions and derive actionable intelligence from these streams. After extracting the features of the image by preprocessing, IMS determines online the classifier to use on the extracted features to make a prediction using the context of the image. A key challenge associated with stream mining is that the prediction accuracy of the classifiers is unknown, since the image source is unknown; thus, these accuracies need to be learned online. Another key challenge of stream mining is that learning can only be done by observing the true label, but this is costly to obtain. To address these challenges, we model the image stream mining problem as an active, online contextual experts problem, where the context of the image is used to guide the classifier selection decision. We develop an active learning algorithm and show that it achieves regret sublinear in the number of images that have been observed so far. To further illustrate and assess the performance of our proposed methods, we apply them to diagnose breast cancer from the images of cellular samples obtained from the fine needle aspirate of breast mass. Our findings show that very high diagnosis accuracy can be achieved by actively obtaining only a small fraction of true labels through surgical biopsies. Other applications include video surveillance and video traffic monitoring.