Cell biology is fundamentally limited in its ability to collect complete data on cellular phenotypes and the wide range of responses to perturbation. Areas such as computer vision and speech recognition have addressed this problem of characterizing unseen or unlabeled conditions with the combined advances of big data, deep learning, and computing resources in the past 5 years. Similarly, recent advances in machine learning approaches enabled by single-cell data start to address prediction tasks in perturbation response modeling. We first define objectives in learning perturbation response in single-cell omics; survey existing approaches, resources, and datasets (https://github.com/theislab/sc-pert); and discuss how a perturbation atlas can enable deep learning models to construct an informative perturbation latent space. We then examine future avenues toward more powerful and explainable modeling using deep neural networks, which enable the integration of disparate information sources and an understanding of heterogeneous, complex, and unseen systems.

Single-cell image analysis provides a powerful approach for studying cell-to-cell heterogeneity, which is an important attribute of isogenic cell populations, from microbial cultures to individual cells in multicellular organisms. This phenotypic variability must be explained at a mechanistic level if biologists are to fully understand cellular function and address the genotype-to-phenotype relationship. Variability in single-cell phenotypes is obscured by bulk readouts or averaging of phenotypes from individual cells in a sample; thus, single-cell image analysis enables a higher resolution view of cellular function. Here, we consider examples of both small- and large-scale studies carried out with isogenic cell populations assessed by fluorescence microscopy, and we illustrate the advantages, challenges, and the promise of quantitative single-cell image analysis.

Molecular translation systems provide a genetically encoded framework for protein synthesis, which is essential for all life. Engineering these systems to incorporate non-canonical amino acids (ncAAs) into peptides and proteins has opened many exciting opportunities in chemical and synthetic biology. Here, we review recent advances that are transforming our ability to engineer molecular translation systems. In cell-based systems, new processes to synthesize recoded genomes, tether ribosomal subunits, and engineer orthogonality with high-throughput workflows have emerged. In cell-free systems, adoption of flexizyme technology and cell-free ribosome synthesis and evolution platforms are expanding the limits of chemistry at the ribosome's RNA-based active site. Looking forward, innovations will deepen understanding of molecular translation and provide a path to polymers with previously unimaginable structures and functions.

A distinctive feature of many biological systems is their ability to adapt to persistent stimuli or disturbances that would otherwise drive them away from a desirable steady state. The resulting stasis enables organisms to function reliably while being subjected to very different external environments. This perspective concerns a stringent type of biological adaptation, robust perfect adaptation (RPA), that is resilient to certain network and parameter perturbations. As in engineered control systems, RPA requires that the regulating network satisfy certain structural constraints that cannot be avoided. We elucidate these ideas using biological examples from systems and synthetic biology. We then argue that understanding the structural constraints underlying RPA allows us to look past implementation details and offers a compelling means to unravel regulatory biological complexity.

We, and all organisms, are an evolutionary masterpiece of multiscale feedback control. Feedback loops enable our cells to grow and then stop at the right size, to divide and self-repair, and to respond with agility to their changing environment. Individual cells engage in long range extracellular feedback with other cells, ensuring continued homeostasis of our tissues and organs. Many long ranging feedback loops regulate vital physiological variables. Here, I will argue that focused efforts to understand the properties and constraints of biological feedback control networks should be central to the quest of understanding life. I will also propose many pressing challenges in this field and review conceptual frameworks that might be consequential for addressing them.

2019 marked the 75th anniversary of the publication of Erwin Schrödinger's What Is Life?, a short book described by Roger Penrose in his preface to a reprint of this classic as "among the most influential scientific writings of the 20th century." In this article, I review the long argument made by Schrödinger as he mused on how the laws of physics could help us understand "the events in space and time which take place within the spatial boundary of a living organism." Though Schrödinger's book is often hailed for its influence on some of the titans who founded molecular biology, this article takes a different tack. Instead of exploring the way the book touched biologists such as James Watson and Francis Crick, as well as its critical reception by others such as Linus Pauling and Max Perutz, I argue that Schrödinger's classic is a timeless manifesto, rather than a dated historical curiosity. What Is Life? is full of timely outlooks and approaches to understanding the mysterious living world that includes and surrounds us and can instead be viewed as a call to arms to tackle the great unanswered challenges in the study of living matter that remain for 21st century science.

Computational and mathematical models are key to obtain a system-level understanding of biological processes, but their limitations have to be clearly defined to allow their proper application and interpretation. Crowdsourced benchmarks in the form of challenges provide an unbiased assessment of methods, and for the past decade, the Dialogue for Reverse Engineering Assessment and Methods (DREAM) organized more than 15 systems biology challenges. From transcription factor binding to dynamical network models, from signaling networks to gene regulation, from whole-cell models to cell-lineage reconstruction, and from single-cell positioning in a tissue to drug combinations and cell survival, the breadth is broad. To celebrate the 5-year anniversary of Cell Systems, we review the genesis of these systems biology challenges and discuss how interlocking the forward- and reverse-modeling paradigms allows to push the rim of systems biology. This approach will persist for systems levels approaches in biology and medicine.

Biological systems are by nature multiscale, consisting of subsystems that factor into progressively smaller units in a deeply hierarchical structure. At any level of the hierarchy, an ever-increasing diversity of technologies can be applied to characterize the corresponding biological units and their relations, resulting in large networks of physical or functional proximities-e.g., proximities of amino acids within a protein, of proteins within a complex, or of cell types within a tissue. Here, we review general concepts and progress in using network proximity measures as a basis for creation of multiscale hierarchical maps of biological systems. We discuss the functionalization of these maps to create predictive models, including those useful in translation of genotype to phenotype, along with strategies for model visualization and challenges faced by multiscale modeling in the near future. Collectively, these approaches enable a unified hierarchical approach to biological data, with application from the molecular to the macroscopic.

The rise of systems biology has ushered a new paradigm: the view of the cell as a system that processes environmental inputs to drive phenotypic outputs. Synthetic biology provides a complementary approach, allowing us to program cell behavior through the addition of synthetic genetic devices into the cellular processor. These devices, and the complex genetic circuits they compose, are engineered using a design-prototype-test cycle, allowing for predictable device performance to be achieved in a context-dependent manner. Within mammalian cells, context effects impact device performance at multiple scales, including the genetic, cellular, and extracellular levels. In order for synthetic genetic devices to achieve predictable behaviors, approaches to overcome context dependence are necessary. Here, we describe control systems approaches for achieving context-aware devices that are robust to context effects. We then consider cell fate programing as a case study to explore the potential impact of context-aware devices for regenerative medicine applications.

Quantitative systems biology, in which predictive mathematical models are constructed to guide the design of experiments and predict experimental outcomes, is at an exciting transition point, where the foundational scientific principles are becoming established, but the impact is not yet global. The next steps necessary for mathematical modeling to transform biological research and applications, in the same way it has already transformed other fields, is not completely clear. The purpose of this perspective is to forecast possible answers to this question-what needs to happen next-by drawing on the experience gained in another field, specifically meteorology. We review here a number of lessons learned in weather prediction that are directly relevant to biological systems modeling, and that we believe can enable the same kinds of global impact in our field as atmospheric modeling makes today.

Although some cell types may be defined anatomically or by physiological function, a rigorous definition of cell state remains elusive. Here, we develop a quantitative, imaging-based platform for the systematic and automated classification of subcellular organization in single cells. We use this platform to quantify subcellular organization and gene expression in >30,000 individual human induced pluripotent stem cell-derived cardiomyocytes, producing a publicly available dataset that describes the population distributions of local and global sarcomere organization, mRNA abundance, and correlations between these traits. While the mRNA abundance of some phenotypically important genes correlates with subcellular organization (e.g., the beta-myosin heavy chain, MYH7), these two cellular metrics are heterogeneous and often uncorrelated, which suggests that gene expression alone is not sufficient to classify cell states. Instead, we posit that cell state should be defined by observing full distributions of quantitative, multidimensional traits in single cells that also account for space, time, and function.

Here, we review extant observations of protein phosphorylation and small-molecule interactions in metabolism and ask which of their specific regulatory functions are conserved in Escherichia coli and Homo sapiens. While the number of phosphosites is dramatically higher in humans, the number of metabolite-protein interactions remains largely constant. Moreover, we found the regulatory logic of metabolite-protein interactions, and in many cases also the effector molecules, to be conserved. Post-translational regulation through phosphorylation does not appear to replace this regulation in human but rather seems to add additional opportunities for fine-tuning and more complex responses. The abundance of metabolite-protein interactions in metabolism, their conserved cross-species abundance, and the apparent conservation of regulatory logic across enormous phylogenetic distance demonstrate their relevance for maintaining cellular homeostasis in these ancient biological processes.

Metal ions are essential for life and represent the second most abundant constituent (after water) of any living cell. While the biological importance of inorganic ions has been appreciated for over a century, we are far from a comprehensive understanding of the functional roles that ions play in cells and organisms. In particular, recent advances are challenging the traditional view that cells maintain constant levels of ion concentrations (ion homeostasis). In fact, the ionic composition (metallome) of cells appears to be purposefully dynamic. The scientific journey that started over 60 years ago with the seminal work by Hodgkin and Huxley on action potentials in neurons is far from reaching its end. New evidence is uncovering how changes in ionic composition regulate unexpected cellular functions and physiology, especially in bacteria, thereby hinting at the evolutionary origins of the dynamic metallome. It is an exciting time for this field of biology, which we discuss and refer to here as IonoBiology.

Folding a linear chain of amino acids into a three-dimensional protein is a complex physical process that ultimately confers an impressive range of diverse functions. Although recent advances have driven significant progress in predicting three-dimensional protein structures from sequence, proteins are not static molecules. Rather, they exist as complex conformational ensembles defined by energy landscapes spanning the space of sequence and conditions. Quantitatively mapping the physical parameters that dictate these landscapes and protein stability is therefore critical to develop models that are capable of predicting how mutations alter function of proteins in disease and informing the design of proteins with desired functions. Here, we review the approaches that are used to quantify protein stability at a variety of scales, from returning multiple thermodynamic and kinetic measurements for a single protein sequence to yielding indirect insights into folding across a vast sequence space. The physical parameters derived from these approaches will provide a foundation for models that extend beyond the structural prediction to capture the complexity of conformational ensembles and, ultimately, their function.

Language models have recently emerged as a powerful machine-learning approach for distilling information from massive protein sequence databases. From readily available sequence data alone, these models discover evolutionary, structural, and functional organization across protein space. Using language models, we can encode amino-acid sequences into distributed vector representations that capture their structural and functional properties, as well as evaluate the evolutionary fitness of sequence variants. We discuss recent advances in protein language modeling and their applications to downstream protein property prediction problems. We then consider how these models can be enriched with prior biological knowledge and introduce an approach for encoding protein structural knowledge into the learned representations. The knowledge distilled by these models allows us to improve downstream function prediction through transfer learning. Deep protein language models are revolutionizing protein biology. They suggest new ways to approach protein and therapeutic design. However, further developments are needed to encode strong biological priors into protein language models and to increase their accessibility to the broader community.