Automated profiling of gene function during embryonic development.

Systematic functional profiling of the gene set that directs embryonic development is an important challenge. To tackle this challenge, we used 4D imaging of C. elegans embryogenesis to capture the effects of 500 gene knockdowns and developed an automated approach to compare developmental phenotypes. The automated approach quantifies features-including germ layer cell numbers, tissue position, and tissue shape-to generate temporal curves whose parameterization yields numerical phenotypic signatures. In conjunction with a new similarity metric that operates across phenotypic space, these signatures enabled the generation of ranked lists of genes predicted to have similar functions, accessible in the PhenoBank web portal, for ∼25% of essential development genes. The approach identified new gene and pathway relationships in cell fate specification and morphogenesis and highlighted the utilization of specialized energy generation pathways during embryogenesis. Collectively, the effort establishes the foundation for comprehensive analysis of the gene set that builds a multicellular organism.

Cross-kingdom expression of synthetic genetic elements promotes discovery of metabolites in the human microbiome.

Small molecules encoded by biosynthetic pathways mediate cross-species interactions and harbor untapped potential, which has provided valuable compounds for medicine and biotechnology. Since studying biosynthetic gene clusters in their native context is often difficult, alternative efforts rely on heterologous expression, which is limited by host-specific metabolic capacity and regulation. Here, we describe a computational-experimental technology to redesign genes and their regulatory regions with hybrid elements for cross-species expression in Gram-negative and -positive bacteria and eukaryotes, decoupling biosynthetic capacity from host-range constraints to activate silenced pathways. These synthetic genetic elements enabled the discovery of a class of microbiome-derived nucleotide metabolites-tyrocitabines-from Lactobacillus iners. Tyrocitabines feature a remarkable orthoester-phosphate, inhibit translational activity, and invoke unexpected biosynthetic machinery, including a class of "Amadori synthases" and "abortive" tRNA synthetases. Our approach establishes a general strategy for the redesign, expression, mobilization, and characterization of genetic elements in diverse organisms and communities.

Restoring the Sense of Touch Using a Sensorimotor Demultiplexing Neural Interface.

Paralyzed muscles can be reanimated following spinal cord injury (SCI) using a brain-computer interface (BCI) to enhance motor function alone. Importantly, the sense of touch is a key component of motor function. Here, we demonstrate that a human participant with a clinically complete SCI can use a BCI to simultaneously reanimate both motor function and the sense of touch, leveraging residual touch signaling from his own hand. In the primary motor cortex (M1), residual subperceptual hand touch signals are simultaneously demultiplexed from ongoing efferent motor intention, enabling intracortically controlled closed-loop sensory feedback. Using the closed-loop demultiplexing BCI almost fully restored the ability to detect object touch and significantly improved several sensorimotor functions. Afferent grip-intensity levels are also decoded from M1, enabling grip reanimation regulated by touch signaling. These results demonstrate that subperceptual neural signals can be decoded from the cortex and transformed into conscious perception, significantly augmenting function.

Hand Knob Area of Premotor Cortex Represents the Whole Body in a Compositional Way.

Decades after the motor homunculus was first proposed, it is still unknown how different body parts are intermixed and interrelated in human motor cortical areas at single-neuron resolution. Using multi-unit recordings, we studied how face, head, arm, and leg movements are represented in the hand knob area of premotor cortex (precentral gyrus) in people with tetraplegia. Contrary to traditional expectations, we found strong representation of all movements and a partially "compositional" neural code that linked together all four limbs. The code consisted of (1) a limb-coding component representing the limb to be moved and (2) a movement-coding component where analogous movements from each limb (e.g., hand grasp and toe curl) were represented similarly. Compositional coding might facilitate skill transfer across limbs, and it provides a useful framework for thinking about how the motor system constructs movement. Finally, we leveraged these results to create a whole-body intracortical brain-computer interface that spreads targets across all limbs.

In Silico Labeling: Predicting Fluorescent Labels in Unlabeled Images.

Microscopy is a central method in life sciences. Many popular methods, such as antibody labeling, are used to add physical fluorescent labels to specific cellular constituents. However, these approaches have significant drawbacks, including inconsistency; limitations in the number of simultaneous labels because of spectral overlap; and necessary perturbations of the experiment, such as fixing the cells, to generate the measurement. Here, we show that a computational machine-learning approach, which we call "in silico labeling" (ISL), reliably predicts some fluorescent labels from transmitted-light images of unlabeled fixed or live biological samples. ISL predicts a range of labels, such as those for nuclei, cell type (e.g., neural), and cell state (e.g., cell death). Because prediction happens in silico, the method is consistent, is not limited by spectral overlap, and does not disturb the experiment. ISL generates biological measurements that would otherwise be problematic or impossible to acquire.

Determinants of Polar versus Nematic Organization in Networks of Dynamic Microtubules and Mitotic Motors.

During cell division, mitotic motors organize microtubules in the bipolar spindle into either polar arrays at the spindle poles or a "nematic" network of aligned microtubules at the spindle center. The reasons for the distinct self-organizing capacities of dynamic microtubules and different motors are not understood. Using in vitro reconstitution experiments and computer simulations, we show that the human mitotic motors kinesin-5 KIF11 and kinesin-14 HSET, despite opposite directionalities, can both organize dynamic microtubules into either polar or nematic networks. We show that in addition to the motor properties the natural asymmetry between microtubule plus- and minus-end growth critically contributes to the organizational potential of the motors. We identify two control parameters that capture system composition and kinetic properties and predict the outcome of microtubule network organization. These results elucidate a fundamental design principle of spindle bipolarity and establish general rules for active filament network organization.

Mapping the Neural Substrates of Behavior.

Assigning behavioral functions to neural structures has long been a central goal in neuroscience and is a necessary first step toward a circuit-level understanding of how the brain generates behavior. Here, we map the neural substrates of locomotion and social behaviors for Drosophila melanogaster using automated machine-vision and machine-learning techniques. From videos of 400,000 flies, we quantified the behavioral effects of activating 2,204 genetically targeted populations of neurons. We combined a novel quantification of anatomy with our behavioral analysis to create brain-behavior correlation maps, which are shared as browsable web pages and interactive software. Based on these maps, we generated hypotheses of regions of the brain causally related to sensory processing, locomotor control, courtship, aggression, and sleep. Our maps directly specify genetic tools to target these regions, which we used to identify a small population of neurons with a role in the control of walking.

Cardiomyocyte orientation recovery at micrometer scale reveals long-axis fiber continuum in heart walls.

Coordinated cardiomyocyte contraction drives the mammalian heart to beat and circulate blood. No consensus model of cardiomyocyte geometrical arrangement exists, due to the limited spatial resolution of whole heart imaging methods and the piecemeal nature of studies based on histological sections. By combining microscopy and computer vision, we produced the first-ever three-dimensional cardiomyocyte orientation reconstruction across mouse ventricular walls at the micrometer scale, representing a gain of three orders of magnitude in spatial resolution. We recovered a cardiomyocyte arrangement aligned to the long-axis direction of the outer ventricular walls. This cellular network lies in a thin shell and forms a continuum with longitudinally arranged cardiomyocytes in the inner walls, with a complex geometry at the apex. Our reconstruction methods can be applied at fine spatial scales to further understanding of heart wall electrical function and mechanics, and set the stage for the study of micron-scale fiber remodeling in heart disease.

Measuring diversity in Hollywood through the large-scale computational analysis of film.

Movies are a massively popular and influential form of media, but their computational study at scale has largely been off-limits to researchers in the United States due to the Digital Millennium Copyright Act. In this work, we illustrate use of a new regulatory framework to enable computational research on film that permits circumvention of technological protection measures on digital video discs (DVDs). We use this exemption to legally digitize a collection of 2,307 films representing the top 50 movies by U.S. box office over the period 1980 to 2022, along with award nominees. We design a computational pipeline for measuring the representation of gender and race/ethnicity in film, drawing on computer vision models for recognizing actors and human perceptions of gender and race/ethnicity. Doing so allows us to learn substantive facts about representation and diversity in Hollywood over this period, confirming earlier studies that see an increase in diversity over the past decade, while allowing us to use computational methods to uncover a range of ad hoc analytical findings. Our work illustrates the affordances of the data-driven analysis of film at a large scale.

Helical coiled nucleosome chromosome architectures during cell cycle progression.

Recent studies showed an interphase chromosome architecture-a specific coiled nucleosome structure-derived from cryopreserved EM tomograms, and dispersed throughout the nucleus. The images were computationally processed to fill in the missing wedges of data caused by incomplete tomographic tilts. The resulting structures increased z-resolution enabling an extension of the proposed architecture to that of mitotic chromosomes. Here, we provide additional insights into the chromosome architecture that was recently published [M. Elbaum et al., Proc. Natl. Acad. Sci. U.S.A. 119, e2119101119 (2022)]. We build on the defined chromosomes time-dependent structures in an effort to probe their dynamics. Variants of the coiled chromosome structures, possibly further defining specific regions, are discussed. We propose, based on generalized specific uncoiling of mitotic chromosomes in telophase, large-scale reorganization of interphase chromosomes. Chromosome territories, organized as micron-sized small patches, are constructed, satisfying complex volume considerations. Finally, we unveiled the structures of replicated coiled chromosomes, still attached to centromeres, as part of chromosome architecture.

Evaluating language models for mathematics through interactions.

There is much excitement about the opportunity to harness the power of large language models (LLMs) when building problem-solving assistants. However, the standard methodology of evaluating LLMs relies on static pairs of inputs and outputs; this is insufficient for making an informed decision about which LLMs are best to use in an interactive setting, and how that varies by setting. Static assessment therefore limits how we understand language model capabilities. We introduce CheckMate, an adaptable prototype platform for humans to interact with and evaluate LLMs. We conduct a study with CheckMate to evaluate three language models (InstructGPT, ChatGPT, and GPT-4) as assistants in proving undergraduate-level mathematics, with a mixed cohort of participants from undergraduate students to professors of mathematics. We release the resulting interaction and rating dataset, MathConverse. By analyzing MathConverse, we derive a taxonomy of human query behaviors and uncover that despite a generally positive correlation, there are notable instances of divergence between correctness and perceived helpfulness in LLM generations, among other findings. Further, we garner a more granular understanding of GPT-4 mathematical problem-solving through a series of case studies, contributed by experienced mathematicians. We conclude with actionable takeaways for ML practitioners and mathematicians: models that communicate uncertainty, respond well to user corrections, and can provide a concise rationale for their recommendations, may constitute better assistants. Humans should inspect LLM output carefully given their current shortcomings and potential for surprising fallibility.

Algorithm-assisted discovery of an intrinsic order among mathematical constants.

In recent decades, a growing number of discoveries in mathematics have been assisted by computer algorithms, primarily for exploring large parameter spaces. As computers become more powerful, an intriguing possibility arises-the interplay between human intuition and computer algorithms can lead to discoveries of mathematical structures that would otherwise remain elusive. Here, we demonstrate computer-assisted discovery of a previously unknown mathematical structure, the conservative matrix field. In the spirit of the Ramanujan Machine project, we developed a massively parallel computer algorithm that found a large number of formulas, in the form of continued fractions, for numerous mathematical constants. The patterns arising from those formulas enabled the construction of the first conservative matrix fields and revealed their overarching properties. Conservative matrix fields unveil unexpected relations between different mathematical constants, such as π and ln(2), or e and the Gompertz constant. The importance of these matrix fields is further realized by their ability to connect formulas that do not have any apparent relation, thus unifying hundreds of existing formulas and generating infinitely many new formulas. We exemplify these implications on values of the Riemann zeta function ζ (n), studied for centuries across mathematics and physics. Matrix fields also enable new mathematical proofs of irrationality. For example, we use them to generalize the celebrated proof by Apéry of the irrationality of ζ (3). Utilizing thousands of personal computers worldwide, our research strategy demonstrates the power of large-scale computational approaches to tackle longstanding open problems and discover unexpected connections across diverse fields of science.

Promising tools for future drug discovery and development in antiarrhythmic therapy.

Arrhythmia refers to irregularities in the rate and rhythm of the heart, with symptoms spanning from mild palpitations to life-threatening arrhythmias and sudden cardiac death (SCD). The complex molecular nature of arrhythmias complicates the selection of appropriate treatment. Current therapies involve the use of antiarrhythmic drugs (class I-IV) with limited efficacy and dangerous side effects and implantable pacemakers and cardioverter-defibrillators with hardware-related complications and inappropriate shocks. The number of novel antiarrhythmic drug in the development pipeline has decreased substantially during the last decade and underscores uncertainties regarding future developments in this field. Consequently, arrhythmia treatment poses significant challenges, prompting the need for alternative approaches. Remarkably, innovative drug discovery and development technologies show promise in helping advance antiarrhythmic therapies. Here, we review unique characteristics and the transformative potential of emerging technologies that offer unprecedented opportunities for transitioning from traditional antiarrhythmics to next-generation therapies. We assess stem cell technology, emphasizing the utility of innovative cell profiling using multi-omics, high-throughput screening, and advanced computational modeling in developing treatments tailored precisely to individual genetic and physiological profiles. We offer insights into gene therapy, peptide and peptibody approaches for drug delivery. We finally discuss potential strengths and weaknesses of such techniques in reducing adverse effects and enhancing overall treatment outcomes, leading to more effective, specific, and safer therapies. Altogether, this comprehensive overview introduces innovative avenues for personalized rhythm therapy, with particular emphasis on drug discovery, aiming to advance the arrhythmia treatment landscape and the prevention of SCD. Significance Statement Arrhythmias and sudden cardiac death account for 15-20% of deaths worldwide. However, current antiarrhythmic therapies are ineffective and with dangerous side effects. Here, we review the field of arrhythmia treatment underscoring the slow progress in advancing the cardiac rhythm therapy pipeline and the uncertainties regarding evolution of this field. We provide information on how emerging technological and experimental tools can help accelerate progress and address the limitations of antiarrhythmic drug discovery.

filoVision - using deep learning and tip markers to automate filopodia analysis.

Filopodia are slender, actin-filled membrane projections used by various cell types for environment exploration. Analyzing filopodia often involves visualizing them using actin, filopodia tip or membrane markers. Due to the diversity of cell types that extend filopodia, from amoeboid to mammalian, it can be challenging for some to find a reliable filopodia analysis workflow suited for their cell type and preferred visualization method. The lack of an automated workflow capable of analyzing amoeboid filopodia with only a filopodia tip label prompted the development of filoVision. filoVision is an adaptable deep learning platform featuring the tools filoTips and filoSkeleton. filoTips labels filopodia tips and the cytosol using a single tip marker, allowing information extraction without actin or membrane markers. In contrast, filoSkeleton combines tip marker signals with actin labeling for a more comprehensive analysis of filopodia shafts in addition to tip protein analysis. The ZeroCostDL4Mic deep learning framework facilitates accessibility and customization for different datasets and cell types, making filoVision a flexible tool for automated analysis of tip-marked filopodia across various cell types and user data.

Image-based molecular representation learning for drug development: a survey.

Artificial intelligence (AI) powered drug development has received remarkable attention in recent years. It addresses the limitations of traditional experimental methods that are costly and time-consuming. While there have been many surveys attempting to summarize related research, they only focus on general AI or specific aspects such as natural language processing and graph neural network. Considering the rapid advance on computer vision, using the molecular image to enable AI appears to be a more intuitive and effective approach since each chemical substance has a unique visual representation. In this paper, we provide the first survey on image-based molecular representation for drug development. The survey proposes a taxonomy based on the learning paradigms in computer vision and reviews a large number of corresponding papers, highlighting the contributions of molecular visual representation in drug development. Besides, we discuss the applications, limitations and future directions in the field. We hope this survey could offer valuable insight into the use of image-based molecular representation learning in the context of drug development.

Protein language models are performant in structure-free virtual screening.

Hitherto virtual screening (VS) has been typically performed using a structure-based drug design paradigm. Such methods typically require the use of molecular docking on high-resolution three-dimensional structures of a target protein-a computationally-intensive and time-consuming exercise. This work demonstrates that by employing protein language models and molecular graphs as inputs to a novel graph-to-transformer cross-attention mechanism, a screening power comparable to state-of-the-art structure-based models can be achieved. The implications thereof include highly expedited VS due to the greatly reduced compute required to run this model, and the ability to perform early stages of computer-aided drug design in the complete absence of 3D protein structures.

Solving independent set problems with photonic quantum circuits.

An independent set (IS) is a set of vertices in a graph such that no edge connects any two vertices. In adiabatic quantum computation [E. Farhi, et al., Science 292, 472-475 (2001); A. Das, B. K. Chakrabarti, Rev. Mod. Phys. 80, 1061-1081 (2008)], a given graph G(V, E) can be naturally mapped onto a many-body Hamiltonian [Formula: see text], with edges [Formula: see text] being the two-body interactions between adjacent vertices [Formula: see text]. Thus, solving the IS problem is equivalent to finding all the computational basis ground states of [Formula: see text]. Very recently, non-Abelian adiabatic mixing (NAAM) has been proposed to address this task, exploiting an emergent non-Abelian gauge symmetry of [Formula: see text] [B. Wu, H. Yu, F. Wilczek, Phys. Rev. A 101, 012318 (2020)]. Here, we solve a representative IS problem [Formula: see text] by simulating the NAAM digitally using a linear optical quantum network, consisting of three C-Phase gates, four deterministic two-qubit gate arrays (DGA), and ten single rotation gates. The maximum IS has been successfully identified with sufficient Trotterization steps and a carefully chosen evolution path. Remarkably, we find IS with a total probability of 0.875(16), among which the nontrivial ones have a considerable weight of about 31.4%. Our experiment demonstrates the potential advantage of NAAM for solving IS-equivalent problems.

Distributed feedforward and feedback cortical processing supports human speech production.

Speech production is a complex human function requiring continuous feedforward commands together with reafferent feedback processing. These processes are carried out by distinct frontal and temporal cortical networks, but the degree and timing of their recruitment and dynamics remain poorly understood. We present a deep learning architecture that translates neural signals recorded directly from the cortex to an interpretable representational space that can reconstruct speech. We leverage learned decoding networks to disentangle feedforward vs. feedback processing. Unlike prevailing models, we find a mixed cortical architecture in which frontal and temporal networks each process both feedforward and feedback information in tandem. We elucidate the timing of feedforward and feedback-related processing by quantifying the derived receptive fields. Our approach provides evidence for a surprisingly mixed cortical architecture of speech circuitry together with decoding advances that have important implications for neural prosthetics.

Urban visual intelligence: Uncovering hidden city profiles with street view images.

A longstanding line of research in urban studies explores how cities can be understood through their appearance. However, what remains unclear is to what extent urban dwellers' everyday life can be explained by the visual clues of the urban environment. In this paper, we address this question by applying a computer vision model to 27 million street view images across 80 counties in the United States. Then, we use the spatial distribution of notable urban features identified through the street view images, such as street furniture, sidewalks, building façades, and vegetation, to predict the socioeconomic profiles of their immediate neighborhood. Our results show that these urban features alone can account for up to 83% of the variance in people's travel behavior, 62% in poverty status, 64% in crime, and 68% in health behaviors. The results outperform models based on points of interest (POI), population, and other demographic data alone. Moreover, incorporating urban features captured from street view images can improve the explanatory power of these other methods by 5% to 25%. We propose "urban visual intelligence" as a process to uncover hidden city profiles, infer, and synthesize urban information with computer vision and street view images. This study serves as a foundation for future urban research interested in this process and understanding the role of visual aspects of the city.

Biophysics of Frequency-Dependent Variation in Paresthesia and Pain Relief during Spinal Cord Stimulation.

The neurophysiological effects of spinal cord stimulation (SCS) for chronic pain are poorly understood, resulting in inefficient failure-prone programming protocols and inadequate pain relief. Nonetheless, novel stimulation patterns are regularly introduced and adopted clinically. Traditionally, paresthetic sensation is considered necessary for pain relief, although novel paradigms provide analgesia without paresthesia. However, like pain relief, the neurophysiological underpinnings of SCS-induced paresthesia are unknown. Here, we paired biophysical modeling with clinical paresthesia thresholds (of both sexes) to investigate how stimulation frequency affects the neural response to SCS relevant to paresthesia and analgesia. Specifically, we modeled the dorsal column (DC) axonal response, dorsal column nucleus (DCN) synaptic transmission, conduction failure within DC fiber collaterals, and dorsal horn network output. Importantly, we found that high-frequency stimulation reduces DC fiber activation thresholds, which in turn accurately predicts clinical paresthesia perception thresholds. Furthermore, we show that high-frequency SCS produces asynchronous DC fiber spiking and ultimately asynchronous DCN output, offering a plausible biophysical basis for why high-frequency SCS is less comfortable and produces qualitatively different sensation than low-frequency stimulation. Finally, we demonstrate that the model dorsal horn network output is sensitive to SCS-inherent variations in spike timing, which could contribute to heterogeneous pain relief across patients. Importantly, we show that model DC fiber collaterals cannot reliably follow high-frequency stimulation, strongly affecting the network output and typically producing antinociceptive effects at high frequencies. Altogether, these findings clarify how SCS affects the nervous system and provide insight into the biophysics of paresthesia generation and pain relief.