Small molecule autoencoders: architecture engineering to optimize latent space utility and sustainability
Ontology highlight
ABSTRACT:
SUBMITTER: Oestreich M
PROVIDER: S-EPMC10913550 | biostudies-literature | 2024 Jan
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:The repertoire of growth factors determines the biological engagement of human mesenchymal stromal cells (hMSCs) in processes such as immunomodulation and tissue repair. Hypoxia is a strong modulator of the secretome and well known stimuli to increase the secretion of pro-angiogenic molecules. In this manuscript, we employed a high throughput screening assay on an hMSCs cell line in order to identify small molecules that mimic hypoxia. Importantly, we show that the effect of these small molecules was cell type/species dependent, but we identified phenanthroline as a robust hit in several cell types. We show that phenanthroline induces high expression of hypoxia-target genes in hMSCs when compared with deferoxamine (DFO) (a.k.a desferrioxamine B, a known hypoxia mimic) and hypoxia incubator (2% O2). Interestingly, our microarray and proteomics analysis show that only phenanthroline induced high expression and secretion of another angiogenic cytokine, interleukin-8, suggesting that the mechanism of phenanthroline-induced hypoxia is distinct from DFO and hypoxia and involves the activation of other signaling pathways. We showed that phenanthroline alone was sufficient to induce blood vessel formation in a Matrigel plug assay in vivo paving the way to its application in ischeamic-related diseases.
2018-10-04 | E-MTAB-7209 | biostudies-arrayexpress
Project description:In the era of data science, data-driven algorithms have emerged as powerful platforms that can consolidate bioisosteric rules for preferential modifications on small molecules with a common molecular scaffold. Here we present complementary data-driven algorithms to minimize the search in chemical space for phenylthiazole-containing molecules that bind the RNA hairpin within the ribosomal peptidyl transferase center (PTC) of Mycobacterium tuberculosis. Our results indicate visual, geometrical, and chemical features that enhance the binding to the targeted RNA. Functional validation was conducted after synthesizing 10 small molecules pinpointed computationally. Four of the 10 were found to be potent inhibitors that target hairpin 91 in the ribosomal PTC of M. tuberculosis and, as a result, stop translation.
| S-EPMC8811966 | biostudies-literature
Project description:Allosteric transcription factors (aTFs) have proven widely applicable for biotechnology and synthetic biology as ligand-specific biosensors enabling real-time monitoring, selection and regulation of cellular metabolism. However, both the biosensor specificity and the correlation between ligand concentration and biosensor output signal, also known as the transfer function, often needs to be optimized before meeting application needs. Here, we present a versatile and high-throughput method to evolve prokaryotic aTF specificity and transfer functions in a eukaryote chassis, namely baker's yeast Saccharomyces cerevisiae. From a single round of mutagenesis of the effector-binding domain (EBD) coupled with various toggled selection regimes, we robustly select aTF variants of the cis,cis-muconic acid-inducible transcription factor BenM evolved for change in ligand specificity, increased dynamic output range, shifts in operational range, and a complete inversion-of-function from activation to repression. Importantly, by targeting only the EBD, the evolved biosensors display DNA-binding affinities similar to BenM, and are functional when ported back into a prokaryotic chassis. The developed platform technology thus leverages aTF evolvability for the development of new host-agnostic biosensors with user-defined small-molecule specificities and transfer functions.
| S-EPMC6943132 | biostudies-literature
Project description:Probabilistic generative models are attractive for scientific modeling because their inferred parameters can be used to generate hypotheses and design experiments. This requires that the learned model provides an accurate representation of the input data and yields a latent space that effectively predicts outcomes relevant to the scientific question. Supervised Variational Autoencoders (SVAEs) have previously been used for this purpose, as a carefully designed decoder can be used as an interpretable generative model of the data, while the supervised objective ensures a predictive latent representation. Unfortunately, the supervised objective forces the encoder to learn a biased approximation to the generative posterior distribution, which renders the generative parameters unreliable when used in scientific models. This issue has remained undetected as reconstruction losses commonly used to evaluate model performance do not detect bias in the encoder. We address this previously-unreported issue by developing a second-order supervision framework (SOS-VAE) that updates the decoder parameters, rather than the encoder, to induce a predictive latent representation. This ensures that the encoder maintains a reliable posterior approximation and the decoder parameters can be effectively interpreted. We extend this technique to allow the user to trade-off the bias in the generative parameters for improved predictive performance, acting as an intermediate option between SVAEs and our new SOS-VAE. We also use this methodology to address missing data issues that often arise when combining recordings from multiple scientific experiments. We demonstrate the effectiveness of these developments using synthetic data and electrophysiological recordings with an emphasis on how our learned representations can be used to design scientific experiments.
| S-EPMC9910304 | biostudies-literature
Project description:Targeting cellular RNA by small molecules has come to the forefront of biotechnology and holds great promise for therapeutic use. Strategies to identify, validate and optimize these molecules are essential, but are still lacking in some aspects. In particular, the site-specific covalent labeling and modification of RNA in living cells poses many challenges. Here, we describe a general structure-guided approach to engineer non-covalent RNA aptamer–ligand complexes into their covalent counterparts using a molecular tether. The key is to modify the native ligand with an electrophilic handle that allows it to react specifically with a guanine at the RNA ligand binding site. We show that site-specific cross-linking between ligand and RNA is achieved in mammalian cells upon transfection of a genetically encoded version of the preQ1-I riboswitch aptamer. Further, we showcase the versatility of the tether by engineering the first covalent fluorescent light-up aptamer (coFLAP) out of the non-covalent Pepper FLAP. The coPepper system maintains strong fluorescence in live-cell imaging even after repeated washing. Thus, any background signal arising from unspecific fluorophore accumulation in the cell can be eliminated. In addition, we generated a bifunctional Pepper ligand containing a second handle for bioorthogonal chemistry to allow for easily traceable and efficient pulldown of the covalently linked target RNA. Finally, we provide evidence for the suitability of this tethering strategy for specific drug targeting. Taken together, our results show that functionalized ligands generated by rational design can cross-link site-specifically with target RNAs in cells, and hence, open up a wide range of applications in RNA biology that require irreversible small molecule binding.
2024-09-30 | GSE271728 | GEO
Project description:The high-dimensional character of most biological systems presents genuine challenges for modeling and prediction. Here we propose a neural network-based approach for dimensionality reduction and analysis of biological gene expression data, using, as a case study, a well-known genetic network in the early Drosophila embryo, the gap gene patterning system. We build an autoencoder compressing the dynamics of spatial gap gene expression into a two-dimensional (2D) latent map. The resulting 2D dynamics suggests an almost linear model, with a small bare set of essential interactions. Maternally defined spatial modes control gap genes positioning, without the classically assumed intricate set of repressive gap gene interactions. This, surprisingly, predicts minimal changes of neighboring gap domains when knocking out gap genes, consistent with previous observations. Latent space geometries in maternal mutants are also consistent with the existence of such spatial modes. Finally, we show how positional information is well defined and interpretable as a polar angle in latent space. Our work illustrates how optimization of small neural networks on medium-sized biological datasets is sufficiently informative to capture essential underlying mechanisms of network function.
| S-EPMC9245618 | biostudies-literature
Project description:Methotrexate (MTX), an inhibitor of dihydrofolate reductase, was tethered to an FKBP12 ligand (SLF), and the resulting bifunctional molecule (MTXSLF) potently inhibits either enzyme but not both simultaneously. MTXSLF is cytotoxic to fibroblasts derived from FKBP12-null mice but is detoxified 40-fold by FKBP12 in wild-type fibroblasts. These studies demonstrate that non-target proteins in an otherwise identical genetic background can be used to predictably regulate the biological activity of synthetic molecules.
| S-EPMC1949043 | biostudies-literature
Project description:This paper proposes a shared synapse architecture for autoencoders (AEs), and implements an AE with the proposed architecture as a digital circuit on a field-programmable gate array (FPGA). In the proposed architecture, the values of the synapse weights are shared between the synapses of an input and a hidden layer, and between the synapses of a hidden and an output layer. This architecture utilizes less of the limited resources of an FPGA than an architecture which does not share the synapse weights, and reduces the amount of synapse modules used by half. For the proposed circuit to be implemented into various types of AEs, it utilizes three kinds of parameters; one to change the number of layers' units, one to change the bit width of an internal value, and a learning rate. By altering a network configuration using these parameters, the proposed architecture can be used to construct a stacked AE. The proposed circuits are logically synthesized, and the number of their resources is determined. Our experimental results show that single and stacked AE circuits utilizing the proposed shared synapse architecture operate as regular AEs and as regular stacked AEs. The scalability of the proposed circuit and the relationship between the bit widths and the learning results are also determined. The clock cycles of the proposed circuits are formulated, and this formula is used to estimate the theoretical performance of the circuit when the circuit is used to construct arbitrary networks.
| S-EPMC5854352 | biostudies-literature
Project description:Small integration time steps limit molecular dynamics (MD) simulations to millisecond time scales. Markov state models (MSMs) and equation-free approaches learn low-dimensional kinetic models from MD simulation data by performing configurational or dynamical coarse-graining of the state space. The learned kinetic models enable the efficient generation of dynamical trajectories over vastly longer time scales than are accessible by MD, but the discretization of configurational space and/or absence of a means to reconstruct molecular configurations precludes the generation of continuous atomistic molecular trajectories. We propose latent space simulators (LSS) to learn kinetic models for continuous atomistic simulation trajectories by training three deep learning networks to (i) learn the slow collective variables of the molecular system, (ii) propagate the system dynamics within this slow latent space, and (iii) generatively reconstruct molecular configurations. We demonstrate the approach in an application to Trp-cage miniprotein to produce novel ultra-long synthetic folding trajectories that accurately reproduce atomistic molecular structure, thermodynamics, and kinetics at six orders of magnitude lower cost than MD. The dramatically lower cost of trajectory generation enables greatly improved sampling and greatly reduced statistical uncertainties in estimated thermodynamic averages and kinetic rates.
| S-EPMC8162036 | biostudies-literature
Project description:The development and implementation of sustainable catalytic technologies is key to delivering our net-zero targets. Here we review how engineered enzymes, with a focus on those developed using directed evolution, can be deployed to improve the sustainability of numerous processes and help to conserve our environment. Efficient and robust biocatalysts have been engineered to capture carbon dioxide (CO2) and have been embedded into new efficient metabolic CO2 fixation pathways. Enzymes have been refined for bioremediation, enhancing their ability to degrade toxic and harmful pollutants. Biocatalytic recycling is gaining momentum, with engineered cutinases and PETases developed for the depolymerization of the abundant plastic, polyethylene terephthalate (PET). Finally, biocatalytic approaches for accessing petroleum-based feedstocks and chemicals are expanding, using optimized enzymes to convert plant biomass into biofuels or other high value products. Through these examples, we hope to illustrate how enzyme engineering and biocatalysis can contribute to the development of cleaner and more efficient chemical industry.
| S-EPMC10952289 | biostudies-literature