Search

Amorphous graphene or amorphous monolayer carbon (AMC) is a family of carbon films that exhibit a surprising sensitivity of electronic conductance to morphology. We combine deep learning-enhanced simulation techniques with percolation theory to analyze three morphologically distinct mesoscale AMCs. Our approach avoids the pitfalls of applying periodic boundary conditions to these fundamentally aperiodic systems or equating crystalline inclusions with conducting sites. We reproduce the previously reported dependence of charge conductance on morphology and explore the limitations of partial morphology descriptors in witnessing conductance properties. Finally, we perform crystallinity analysis of conductance networks along the electronic energy spectrum and show that they metamorphose from being localized on crystallites at band edges to localized on defects around the Fermi energy opening the possibility of control through gate voltage.

Amorphous solids form an enormous and underutilized class of materials. In order to drive the discovery of new useful amorphous materials further we need to achieve a closer convergence between computational and experimental methods. In this review, we highlight some of the important gaps between computational simulations and experiments, discuss popular state-of-the-art computational techniques such as the Activation Relaxation Technique nouveau (ARTn) and Reverse Monte Carlo (RMC), and introduce more recent advances: machine learning interatomic potentials (MLIPs) and generative machine learning for simulations of amorphous matter, e.g., the Morphological Autoregressive Protocol (MAP). Examples are drawn from the amorphous silicon and silica literature as well as from molecular glasses. Our outlook stresses the need for new computational methods to extend the time- and length- scales accessible through numerical simulations.

Signed Particle Monte Carlo (SPMC) approach has been used in the past to model steady-state and transient dynamics of the Wigner quasi-distribution for electrons in low dimensional semiconductors. Here we make a step towards high-dimensional quantum phase-space simulation in chemically relevant scenarios by improving the stability and memory demands of SPMC in 2D. We do so by using an unbiased propagator for SPMC to improve trajectory stability and by applying machine learning to reduce memory demands for storage and manipulation of the Wigner potential. We perform computational experiments on a 2D double-well toymodel of proton transfer and demonstrate stable pico-second-long trajectories that require only a modest computational effort.

Disordered molecular systems such as amorphous catalysts, organic thin films, electrolyte solutions, and water are at the cutting edge of computational exploration today. Traditional simulations of such systems at length-scales relevant to experiments in practice require a compromise between model accuracy and quality of sampling. To remedy the situation, we have developed an approach based on generative machine learning called the Morphological Autoregressive Protocol (MAP) which provides computational access to mesoscale disordered molecular configurations at linear cost at generation for materials in which structural correlations decay sufficiently rapidly. The algorithm is implemented using an augmented PixelCNN deep learning architecture that we previously demonstrated produces excellent results in 2 dimensions (2D) for mono-elemental molecular systems. Here, we extend our implementation to multielemental 3D and demonstrate performance using water as our test system in two scenarios: 1. liquid water, and 2. a sample conditioned on the presence of a rare motif. We trained the model on small-scale samples of liquid water produced using path-integral molecular dynamics simulation including nuclear quantum effects under ambient conditions. MAP-generated water configurations are shown to accurately reproduce the properties of the training set and to produce stable trajectories when used as initial conditions in classical and quantum dynamical simulations. We expect our approach to perform equally well on other disordered molecular systems while offering unique advantages in situations when the disorder is quenched rather than equilibrated.

Inverse design of short single-stranded RNA and DNA sequences (aptamers) is the task of finding sequences that satisfy a set of desired criteria. Relevant criteria may be, for example, the presence of specific folding motifs, binding to molecular ligands, sensing properties, etc. Most practical approaches to aptamer design identify a small set of promising candidate sequences using high-throughput experiments (e.g. SELEX), and then optimize performance by introducing only minor modifications to the empirically found candidates. Sequences that possess the desired properties but differ drastically in chemical composition will add diversity to the search space and facilitate the discovery of useful nucleic acid aptamers. Systematic diversification protocols are needed. Here we propose to use an unsupervised machine learning model known as the Potts model to discover new, useful sequences with controllable sequence diversity. We start by training a Potts model using the maximum entropy principle on a small set of empirically identified sequences unified by a common feature. To generate new candidate sequences with a controllable degree of diversity, we take advantage of the model's spectral feature: an energy bandgap separating sequences that are similar to the training set from those that are distinct. By controlling the Potts energy range that is sampled, we generate sequences that are distinct from the training set yet still likely to have the encoded features. To demonstrate performance, we apply our approach to design diverse pools of sequences with specified secondary structure motifs in 30-mer RNA and DNA aptamers.

Design of de novo biological sequences with desired properties, like protein and DNA sequences, often involves an active loop with several rounds of molecule ideation and expensive wet-lab evaluations. These experiments can consist of multiple stages, with increasing levels of precision and cost of evaluation, where candidates are filtered. This makes the diversity of proposed candidates a key consideration in the ideation phase. In this work, we propose an active learning algorithm leveraging epistemic uncertainty estimation and the recently proposed GFlowNets as a generator of diverse candidate solutions, with the objective to obtain a diverse batch of useful (as defined by some utility function, for example, the predicted anti-microbial activity of a peptide) and informative candidates after each round. We also propose a scheme to incorporate existing labeled datasets of candidates, in addition to a reward function, to speed up learning in GFlowNets. We present empirical results on several biological sequence design tasks, and we find that our method generates more diverse and novel batches with high scoring candidates compared to existing approaches., Comment: ICML 2022. 15 pages, 3 figures. Code available at: https://github.com/MJ10/BioSeq-GFN-AL. Updated GFP results