Explore Protein Conformational Space With Variational Autoencoder.
Ontology highlight
ABSTRACT:
SUBMITTER: Tian H
PROVIDER: S-EPMC8633506 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Mapping the ensemble of protein conformations that contribute to function and can be targeted by small molecule drugs remains an outstanding challenge. Here, we explore the use of variational autoencoders for reducing the challenge of dimensionality in the protein structure ensemble generation problem. We convert high-dimensional protein structural data into a continuous, low-dimensional representation, carry out a search in this space guided by a structure quality metric, and then use RoseTTAFold guided by the sampled structural information to generate 3D structures. We use this approach to generate ensembles for the cancer relevant protein K-Ras, train the VAE on a subset of the available K-Ras crystal structures and MD simulation snapshots, and assess the extent of sampling close to crystal structures withheld from training. We find that our latent space sampling procedure rapidly generates ensembles with high structural quality and is able to sample within 1 Å of held-out crystal structures, with a consistency higher than that of MD simulation or AlphaFold2 prediction. The sampled structures sufficiently recapitulate the cryptic pockets in the held-out K-Ras structures to allow for small molecule docking.
| S-EPMC11008089 | biostudies-literature
Project description:A variational autoencoder (VAE) is a machine learning algorithm, useful for generating a compressed and interpretable latent space. These representations have been generated from various biomedical data types and can be used to produce realistic-looking simulated data. However, standard vanilla VAEs suffer from entangled and uninformative latent spaces, which can be mitigated using other types of VAEs such as β-VAE and MMD-VAE. In this project, we evaluated the ability of VAEs to learn cell morphology characteristics derived from cell images. We trained and evaluated these three VAE variants-Vanilla VAE, β-VAE, and MMD-VAE-on cell morphology readouts and explored the generative capacity of each model to predict compound polypharmacology (the interactions of a drug with more than one target) using an approach called latent space arithmetic (LSA). To test the generalizability of the strategy, we also trained these VAEs using gene expression data of the same compound perturbations and found that gene expression provides complementary information. We found that the β-VAE and MMD-VAE disentangle morphology signals and reveal a more interpretable latent space. We reliably simulated morphology and gene expression readouts from certain compounds thereby predicting cell states perturbed with compounds of known polypharmacology. Inferring cell state for specific drug mechanisms could aid researchers in developing and identifying targeted therapeutics and categorizing off-target effects in the future.
| S-EPMC8906577 | biostudies-literature
Project description:The structural diversity of chemical libraries, which are systematic collections of compounds that have potential to bind to biomolecules, can be represented by chemical latent space. A chemical latent space is a projection of a compound structure into a mathematical space based on several molecular features, and it can express structural diversity within a compound library in order to explore a broader chemical space and generate novel compound structures for drug candidates. In this study, we developed a deep-learning method, called NP-VAE (Natural Product-oriented Variational Autoencoder), based on variational autoencoder for managing hard-to-analyze datasets from DrugBank and large molecular structures such as natural compounds with chirality, an essential factor in the 3D complexity of compounds. NP-VAE was successful in constructing the chemical latent space from large-sized compounds that were unable to be handled in existing methods, achieving higher reconstruction accuracy, and demonstrating stable performance as a generative model across various indices. Furthermore, by exploring the acquired latent space, we succeeded in comprehensively analyzing a compound library containing natural compounds and generating novel compound structures with optimized functions.
| S-EPMC10654724 | biostudies-literature
Project description:BACKGROUND:The sequence of nucleotides in an RNA determines the possible base pairs for an RNA fold and thus also determines the overall shape and function of an RNA. The Swellix program presented here combines a helix abstraction with a combinatorial approach to the RNA folding problem in order to compute all possible non-pseudoknotted RNA structures for RNA sequences. The Swellix program builds on the Crumple program and can include experimental constraints on global RNA structures such as the minimum number and lengths of helices from crystallography, cryoelectron microscopy, or in vivo crosslinking and chemical probing methods. RESULTS:The conceptual advance in Swellix is to count helices and generate all possible combinations of helices rather than counting and combining base pairs. Swellix bundles similar helices and includes improvements in memory use and efficient parallelization. Biological applications of Swellix are demonstrated by computing the reduction in conformational space and entropy due to naturally modified nucleotides in tRNA sequences and by motif searches in Human Endogenous Retroviral (HERV) RNA sequences. The Swellix motif search reveals occurrences of protein and drug binding motifs in the HERV RNA ensemble that do not occur in minimum free energy or centroid predicted structures. CONCLUSIONS:Swellix presents significant improvements over Crumple in terms of efficiency and memory use. The efficient parallelization of Swellix enables the computation of sequences as long as 418 nucleotides with sufficient experimental constraints. Thus, Swellix provides a practical alternative to free energy minimization tools when multiple structures, kinetically determined structures, or complex RNA-RNA and RNA-protein interactions are present in an RNA folding problem.
| S-EPMC5697422 | biostudies-literature
Project description:Over millennia, natural evolution has allowed for the emergence of countless biomolecules with highly specific roles within natural systems. As seen with peptides and proteins, often evolution produces molecules with a similar function but with variable amino acid composition and structure but diverging from a common ancestor, which can limit sequence diversity. Using antimicrobial peptides as a model biomolecule, we train a generative deep learning algorithm on a database of known antimicrobial peptides to generate novel peptide sequences with antimicrobial activity. Using a variational autoencoder, we are able to generate a latent space plot that can be surveyed for peptides with known properties and interpolated across a predictive vector between two defined points to identify novel peptides that show dose-responsive antimicrobial activity. These proof-of-concept studies demonstrate the potential for artificial intelligence-directed methods to generate new antimicrobial peptides and motivate their potential application toward peptide and protein design without the need for exhaustive screening of sequence libraries.
| S-EPMC7450509 | biostudies-literature
Project description:BackgroundMultiple studies have shown the utility of transcriptome-wide RNA-seq profiles as features for machine learning-based prediction of response to chemotherapy in cancer. While tumor transcriptome profiles are publicly available for thousands of tumors for many cancer types, a relatively modest number of tumor profiles are clinically annotated for response to chemotherapy. The paucity of labeled examples and the high dimension of the feature data limit performance for predicting therapeutic response using fully-supervised classification methods. Recently, multiple studies have established the utility of a deep neural network approach, the variational autoencoder (VAE), for generating meaningful latent features from original data. Here, we report the first study of a semi-supervised approach using VAE-encoded tumor transcriptome features and regularized gradient boosted decision trees (XGBoost) to predict chemotherapy drug response for five cancer types: colon, pancreatic, bladder, breast, and sarcoma.ResultsWe found: (1) VAE-encoding of the tumor transcriptome preserves the cancer type identity of the tumor, suggesting preservation of biologically relevant information; and (2) as a feature-set for supervised classification to predict response-to-chemotherapy, the unsupervised VAE encoding of the tumor's gene expression profile leads to better area under the receiver operating characteristic curve and area under the precision-recall curve classification performance than the original gene expression profile or the PCA principal components or the ICA components of the gene expression profile, in four out of five cancer types that we tested.ConclusionsGiven high-dimensional "omics" data, the VAE is a powerful tool for obtaining a nonlinear low-dimensional embedding; it yields features that retain biological patterns that distinguish between different types of cancer and that enable more accurate tumor transcriptome-based prediction of response to chemotherapy than would be possible using the original data or their principal components.
| S-EPMC8456615 | biostudies-literature
Project description:Predicting future brain topography can give insight into neural correlates of aging and neurodegeneration. Due to variability in the aging process, it has been challenging to precisely estimate brain topographical change according to aging. Here, we predict age-related brain metabolic change by generating future brain 18F-Fluorodeoxyglucose PET. A cross-sectional PET dataset of cognitively normal subjects with different age was used to develop a generative model. The model generated PET images using age information and characteristic individual features. Predicted regional metabolic changes were correlated with the real changes obtained by follow-up data. This model was applied to produce a brain metabolism aging movie by generating PET at different ages. Normal population distribution of brain metabolic topography at each age was estimated as well. In addition, a generative model using APOE4 status as well as age as inputs revealed a significant effect of APOE4 status on age-related metabolic changes particularly in the calcarine, lingual cortex, hippocampus, and amygdala. It suggested APOE4 could be a factor affecting individual variability in age-related metabolic degeneration in normal elderly. This predictive model may not only be extended to understanding the cognitive aging process, but apply to the development of a preclinical biomarker for various brain disorders.
| S-EPMC6052253 | biostudies-literature
Project description:Resting state functional magnetic resonance imaging (rsfMRI) data exhibits complex but structured patterns. However, the underlying origins are unclear and entangled in rsfMRI data. Here we establish a variational auto-encoder, as a generative model trainable with unsupervised learning, to disentangle the unknown sources of rsfMRI activity. After being trained with large data from the Human Connectome Project, the model has learned to represent and generate patterns of cortical activity and connectivity using latent variables. The latent representation and its trajectory represent the spatiotemporal characteristics of rsfMRI activity. The latent variables reflect the principal gradients of the latent trajectory and drive activity changes in cortical networks. Representational geometry captured as covariance or correlation between latent variables, rather than cortical connectivity, can be used as a more reliable feature to accurately identify subjects from a large group, even if only a short period of data is available in each subject. Our results demonstrate that VAE is a valuable addition to existing tools, particularly suited for unsupervised representation learning of resting state fMRI activity.
| S-EPMC8485214 | biostudies-literature
Project description:An accurate personality model is crucial to many research fields. Most personality models have been constructed using linear factor analysis (LFA). In this paper, we investigate if an effective deep learning tool for factor extraction, the Variational Autoencoder (VAE), can be applied to explore the factor structure of a set of personality variables. To compare VAE with LFA, we applied VAE to an International Personality Item Pool (IPIP) Big 5 dataset and an IPIP HEXACO (Humility-Honesty, Emotionality, Extroversion, Agreeableness, Conscientiousness, Openness) dataset. We found that LFA tends to break factors into ever smaller, yet still significant fractions, when the number of assumed latent factors increases, leading to the need to organize personality variables at the factor level and then the facet level. On the other hand, the factor structure returned by VAE is very stable and VAE only adds noise-like factors after significant factors are found as the number of assumed latent factors increases. VAE reported more stable factors by elevating some facets in the HEXACO scale to the factor level. Since this is a data-driven process that exhausts all stable and significant factors that can be found, it is not necessary to further conduct facet level analysis and it is anticipated that VAE will have broad applications in exploratory factor analysis in personality research.
| S-EPMC9388855 | biostudies-literature
Project description:The clock drawing test (CDT) is an inexpensive tool to screen for dementia. In this study, we examined if a variational autoencoder (VAE) with only two latent variables can capture and encode clock drawing anomalies from a large dataset of unannotated CDTs (n = 13,580) using self-supervised pre-training and use them to classify dementia CDTs (n = 18) from non-dementia CDTs (n = 20). The model was independently validated using a larger cohort consisting of 41 dementia and 50 non-dementia clocks. The classification model built with the parsimonious VAE latent space adequately classified dementia from non-dementia (0.78 area under receiver operating characteristics (AUROC) in the original test dataset and 0.77 AUROC in the secondary validation dataset). The VAE-identified atypical clock features were then reviewed by domain experts and compared with existing literature on clock drawing errors. This study shows that a very small number of latent variables are sufficient to encode important clock drawing anomalies that are predictive of dementia.
| S-EPMC9107463 | biostudies-literature