A deep learning approach to programmable RNA switches

Detection of pks Island mRNAs Using Toehold Sensors in Escherichia coli

Synthetic biologists have applied biomolecular engineering approaches toward the goal of novel biological devices and have shown progress in diverse areas of medicine and biotechnology. Especially promising is the application of synthetic biological devices towards a novel class of molecular diagnostics. As an example, a de-novo-designed riboregulator called toehold switch, with its programmability and compatibility with field-deployable devices showed promising in vitro applications for viral RNA detection such as Zika and Corona viruses. However, the in vivo application of high-performance RNA sensors remains challenging due to the secondary structure of long mRNA species. Here, we introduced ‘Helper RNAs’ that can enhance the functionality of toehold switch sensors by mitigating the effect of secondary structures around a target site. By employing the helper RNAs, previously reported mCherry mRNA sensor showed improved fold-changes in vivo. To further generalize the Helper RNA approaches, we employed automatic design pipeline for toehold sensors that target the essential genes within the pks island, an important target of biomedical research in connection with colorectal cancer. The toehold switch sensors showed fold-changes upon the expression of full-length mRNAs that apparently depended sensitively on the identity of the gene as well as the predicted local structure within the target region of the mRNA. Still, the helper RNAs could improve the performance of toehold switch sensors in many instances, with up to 10-fold improvement over no helper cases. These results suggest that the helper RNA approaches can further assist the design of functional RNA devices in vivo with the aid of the streamlined automatic design software developed here. Further, our solutions for screening and stabilizing single-stranded region of mRNA may find use in other in vivo mRNA-sensing applications such as cas13 crRNA design, transcriptome engineering, and trans-cleaving ribozymes.

Prediction of Breast Cancer Recurrence Using a Deep Convolutional Neural Network Without Region-of-Interest Labeling

Purpose

The present study aimed to assign a risk score for breast cancer recurrence based on pathological whole slide images (WSIs) using a deep learning model.

Methods

A total of 233 WSIs from 138 breast cancer patients were assigned either a low-risk or a high-risk score based on a 70-gene signature. These images were processed into patches of 512x512 pixels by the PyHIST tool and underwent color normalization using the Macenko method. Afterward, out of focus and pixelated patches were removed using the Laplacian algorithm. Finally, the remaining patches (n=294,562) were split into 3 parts for model training (50%), validation (7%) and testing (43%). We used 6 pretrained models for transfer learning and evaluated their performance using accuracy, precision, recall, F1 score, confusion matrix, and AUC. Additionally, to demonstrate the robustness of the final model and its generalization capacity, the testing set was used for model evaluation. Finally, the GRAD-CAM algorithm was used for model visualization.

Results

Six models, namely VGG16, ResNet50, ResNet101, Inception_ResNet, EfficientB5, and Xception, achieved high performance in the validation set with an overall accuracy of 0.84, 0.85, 0.83, 0.84, 0.87, and 0.91, respectively. We selected Xception for assessment of the testing set, and this model achieved an overall accuracy of 0.87 with a patch-wise approach and 0.90 and 1.00 with a patient-wise approach for high-risk and low-risk groups, respectively.

Conclusions

Our study demonstrated the feasibility and high performance of artificial intelligence models trained without region-of-interest labeling for predicting cancer recurrence based on a 70-gene signature risk score.

Programming living sensors for environment, health and biomanufacturing

Synthetic biology offers new tools and capabilities of engineering cells with desired functions for example as new biosensing platforms leveraging engineered microbes. In the last two decades, bacterial cells have been programmed to sense and respond to various input cues for versatile purposes including environmental monitoring, disease diagnosis and adaptive biomanufacturing. Despite demonstrated proof-of-concept success in the laboratory, the real-world applications of microbial sensors have been restricted due to certain technical and societal limitations. Yet, most limitations can be addressed by new technological developments in synthetic biology such as circuit design, biocontainment and machine learning. Here, we summarize the latest advances in synthetic biology and discuss how they could accelerate the development, enhance the performance and address the present limitations of microbial sensors to facilitate their use in the field. We view that programmable living sensors are promising sensing platforms to achieve sustainable, affordable and easy-to-use on-site detection in diverse settings.

Importance of the 5' regulatory region to bacterial synthetic biology applications

The field of synthetic biology is evolving at a fast pace. It is advancing beyond single-gene alterations in single hosts to the logical design of complex circuits and the development of integrated synthetic genomes. Recent breakthroughs in deep learning, which is increasingly used in de novo assembly of DNA components with predictable effects, are also aiding the discipline. Despite advances in computing, the field is still reliant on the availability of pre-characterized DNA parts, whether natural or synthetic, to regulate gene expression in bacteria and make valuable compounds. In this review, we discuss the different bacterial synthetic biology methodologies employed in the creation of 5' regulatory regions - promoters, untranslated regions and 5'-end of coding sequences. We summarize methodologies and discuss their significance for each of the functional DNA components, and highlight the key advances made in bacterial engineering by concentrating on their flaws and strengths. We end the review by outlining the issues that the discipline may face in the near future.

Context-Dependent Stability and Robustness of Genetic Toggle Switches with Leaky Promoters

Multistable switches are ubiquitous building blocks in both systems and synthetic biology. Given their central role, it is thus imperative to understand how their fundamental properties depend not only on the tunable biophysical properties of the switches themselves, but also on their genetic context. To this end, we reveal in this article how these factors shape the essential characteristics of toggle switches implemented using leaky promoters such as their stability and robustness to noise, both at single-cell and population levels. In particular, our results expose the roles that competition for scarce transcriptional and translational resources, promoter leakiness, and cell-to-cell heterogeneity collectively play. For instance, the interplay between protein expression from leaky promoters and the associated cost of relying on shared cellular resources can give rise to tristable dynamics even in the absence of positive feedback. Similarly, we demonstrate that while promoter leakiness always acts against multistability, resource competition can be leveraged to counteract this undesirable phenomenon. Underpinned by a mechanistic model, our results thus enable the context-aware rational design of multistable genetic switches that are directly translatable to experimental considerations, and can be further leveraged during the synthesis of large-scale genetic systems using computer-aided biodesign automation platforms.

In silico, in vitro, and in vivo machine learning in synthetic biology and metabolic engineering

Among the main learning methods reviewed in this study and used in synthetic biology and metabolic engineering are supervised learning, reinforcement and active learning, and in vitro or in vivo learning. In the context of biosynthesis, supervised machine learning is being exploited to predict biological sequence activities, predict structures and engineer sequences, and optimize culture conditions. Active and reinforcement learning methods use training sets acquired through an iterative process generally involving experimental measurements. They are applied to design, engineer, and optimize metabolic pathways and bioprocesses. The nascent but promising developments with in vitro and in vivo learning comprise molecular circuits performing simple tasks such as pattern recognition and classification.

CRISPR-based diagnostics

The accurate and timely diagnosis of disease is a prerequisite for efficient therapeutic intervention and epidemiological surveillance. Diagnostics based on the detection of nucleic acids are among the most sensitive and specific, yet most such assays require costly equipment and trained personnel. Recent developments in diagnostic technologies, in particular those leveraging clustered regularly interspaced short palindromic repeats (CRISPR), aim to enable accurate testing at home, at the point of care and in the field. In this Review, we provide a rundown of the rapidly expanding toolbox for CRISPR-based diagnostics, in particular the various assays, preamplification strategies and readouts, and highlight their main applications in the sensing of a wide range of molecular targets relevant to human health.

Synthetic biology as driver for the biologization of materials sciences

Materials in nature have fascinating properties that serve as a continuous source of inspiration for materials scientists. Accordingly, bio-mimetic and bio-inspired approaches have yielded remarkable structural and functional materials for a plethora of applications. Despite these advances, many properties of natural materials remain challenging or yet impossible to incorporate into synthetic materials. Natural materials are produced by living cells, which sense and process environmental cues and conditions by means of signaling and genetic programs, thereby controlling the biosynthesis, remodeling, functionalization, or degradation of the natural material. In this context, synthetic biology offers unique opportunities in materials sciences by providing direct access to the rational engineering of how a cell senses and processes environmental information and translates them into the properties and functions of materials. Here, we identify and review two main directions by which synthetic biology can be harnessed to provide new impulses for the biologization of the materials sciences: first, the engineering of cells to produce precursors for the subsequent synthesis of materials. This includes materials that are otherwise produced from petrochemical resources, but also materials where the bio-produced substances contribute unique properties and functions not existing in traditional materials. Second, engineered living materials that are formed or assembled by cells or in which cells contribute specific functions while remaining an integral part of the living composite material. We finally provide a perspective of future scientific directions of this promising area of research and discuss science policy that would be required to support research and development in this field.

Representation learning applications in biological sequence analysis

Although remarkable advances have been reported in high-throughput sequencing, the ability to aptly analyze a substantial amount of rapidly generated biological (DNA/RNA/protein) sequencing data remains a critical hurdle. To tackle this issue, the application of natural language processing (NLP) to biological sequence analysis has received increased attention. In this method, biological sequences are regarded as sentences while the single nucleic acids/amino acids or k-mers in these sequences represent the words. Embedding is an essential step in NLP, which performs the conversion of these words into vectors. Specifically, representation learning is an approach used for this transformation process, which can be applied to biological sequences. Vectorized biological sequences can then be applied for function and structure estimation, or as input for other probabilistic models. Considering the importance and growing trend for the application of representation learning to biological research, in the present study, we have reviewed the existing knowledge in representation learning for biological sequence analysis.

MoiRNAiFold: a novel tool for complex in silico RNA design

Novel tools for in silico design of RNA constructs such as riboregulators are required in order to reduce time and cost to production for the development of diagnostic and therapeutic advances. Here, we present MoiRNAiFold, a versatile and user-friendly tool for de novo synthetic RNA design. MoiRNAiFold is based on Constraint Programming and it includes novel variable types, heuristics and restart strategies for Large Neighborhood Search. Moreover, this software can handle dozens of design constraints and quality measures and improves features for RNA regulation control of gene expression, such as Translation Efficiency calculation. We demonstrate that MoiRNAiFold outperforms any previous software in benchmarking structural RNA puzzles from EteRNA. Importantly, with regard to biologically relevant RNA designs, we focus on RNA riboregulators, demonstrating that the designed RNA sequences are functional both in vitro and in vivo. Overall, we have generated a powerful tool for de novo complex RNA design that we make freely available as a web server (https://moiraibiodesign.com/design/).

Genetic toolkits for engineering Rhodococcus species with versatile applications

Rhodococcus spp. are a group of non-model gram-positive bacteria with diverse catabolic activities and strong adaptive capabilities, which enable their wide application in whole-cell biocatalysis, environmental bioremediation, and lignocellulosic biomass conversion. Compared with model microorganisms, the engineering of Rhodococcus is challenging because of the lack of universal molecular tools, high genome GC content (61% ~ 71%), and low transformation and recombination efficiencies. Nevertheless, because of the high interest in Rhodococcus species for bioproduction, various genetic elements and engineering tools have been recently developed for Rhodococcus spp., including R. opacus, R. jostii, R. ruber, and R. erythropolis, leading to the expansion of the genetic toolkits for Rhodococcus engineering. In this article, we provide a comprehensive review of the important developed genetic elements for Rhodococcus, including shuttle vectors, promoters, antibiotic markers, ribosome binding sites, and reporter genes. In addition, we also summarize gene transfer techniques and strategies to improve transformation efficiency, as well as random and precise genome editing tools available for Rhodococcus, including transposition, homologous recombination, recombineering, and CRISPR/Cas9. We conclude by discussing future trends in Rhodococcus engineering. We expect that more synthetic and systems biology tools (such as multiplex genome editing, dynamic regulation, and genome-scale metabolic models) will be adapted and optimized for Rhodococcus.

Sequence-independent RNA sensing and DNA targeting by a split domain CRISPR-Cas12a gRNA switch

CRISPR technologies increasingly require spatiotemporal and dosage control of nuclease activity. One promising strategy involves linking nuclease activity to a cell's transcriptional state by engineering guide RNAs (gRNAs) to function only after complexing with a ‘trigger’ RNA. However, standard gRNA switch designs do not allow independent selection of trigger and guide sequences, limiting gRNA switch application. Here, we demonstrate the modular design of Cas12a gRNA switches that decouples selection of these sequences. The 5′ end of the Cas12a gRNA is fused to two distinct and non-overlapping domains: one base pairs with the gRNA repeat, blocking formation of a hairpin required for Cas12a recognition; the other hybridizes to the RNA trigger, stimulating refolding of the gRNA repeat and subsequent gRNA-dependent Cas12a activity. Using a cell-free transcription-translation system and Escherichia coli, we show that designed gRNA switches can respond to different triggers and target different DNA sequences. Modulating the length and composition of the sensory domain altered gRNA switch performance. Finally, gRNA switches could be designed to sense endogenous RNAs expressed only under specific growth conditions, rendering Cas12a targeting activity dependent on cellular metabolism and stress. Our design framework thus further enables tethering of CRISPR activities to cellular states.

Stability and Robustness of Unbalanced Genetic Toggle Switches in the Presence of Scarce Resources

While the vision of synthetic biology is to create complex genetic systems in a rational fashion, system-level behaviors are often perplexing due to the context-dependent dynamics of modules. One major source of context-dependence emerges due to the limited availability of shared resources, coupling the behavior of disconnected components. Motivated by the ubiquitous role of toggle switches in genetic circuits ranging from controlling cell fate differentiation to optimizing cellular performance, here we reveal how their fundamental dynamic properties are affected by competition for scarce resources. Combining a mechanistic model with nullcline-based stability analysis and potential landscape-based robustness analysis, we uncover not only the detrimental impacts of resource competition, but also how the unbalancedness of the switch further exacerbates them. While in general both of these factors undermine the performance of the switch (by pushing the dynamics toward monostability and increased sensitivity to noise), we also demonstrate that some of the unwanted effects can be alleviated by strategically optimized resource competition. Our results provide explicit guidelines for the context-aware rational design of toggle switches to mitigate our reliance on lengthy and expensive trial-and-error processes, and can be seamlessly integrated into the computer-aided synthesis of complex genetic systems.

Synthetic biology in the clinic: engineering vaccines, diagnostics, and therapeutics

Synthetic biology is a design-driven discipline centered on engineering novel biological functions through the discovery, characterization, and repurposing of molecular parts. Several synthetic biological solutions to critical biomedical problems are on the verge of widespread adoption and demonstrate the burgeoning maturation of the field. Here, we highlight applications of synthetic biology in vaccine development, molecular diagnostics, and cell-based therapeutics, emphasizing technologies approved for clinical use or in active clinical trials. We conclude by drawing attention to recent innovations in synthetic biology that are likely to have a significant impact on future applications in biomedicine.

Sequence-to-function deep learning frameworks for engineered riboregulators

While synthetic biology has revolutionized our approaches to medicine, agriculture, and energy, the design of completely novel biological circuit components beyond naturally-derived templates remains challenging due to poorly understood design rules. Toehold switches, which are programmable nucleic acid sensors, face an analogous design bottleneck; our limited understanding of how sequence impacts functionality often necessitates expensive, time-consuming screens to identify effective switches. Here, we introduce Sequence-based Toehold Optimization and Redesign Model (STORM) and Nucleic-Acid Speech (NuSpeak), two orthogonal and synergistic deep learning architectures to characterize and optimize toeholds. Applying techniques from computer vision and natural language processing, we 'un-box' our models using convolutional filters, attention maps, and in silico mutagenesis. Through transfer-learning, we redesign sub-optimal toehold sensors, even with sparse training data, experimentally validating their improved performance. This work provides sequence-to-function deep learning frameworks for toehold selection and design, augmenting our ability to construct potent biological circuit components and precision diagnostics.