Search

Few existing efforts to predict childhood obesity have included risk factors across the prenatal and early infancy periods, despite evidence that the first 1000 days is critical for obesity prevention. In this study, we employed machine learning techniques to understand the influence of factors in the first 1000 days on body mass index (BMI) values during childhood. We used LASSO regression to identify 13 features in addition to historical weight, height, and BMI that were relevant to childhood obesity. We then developed prediction models based on support vector regression with fivefold cross validation, estimating BMI for three time periods: 30-36 (N = 4204), 36-42 (N = 4130), and 42-48 (N = 2880) months. Our models were developed using 80% of the patients from each period. When tested on the remaining 20% of the patients, the models predicted children's BMI with high accuracy (mean average error [standard deviation] = 0.96[0.02] at 30-36 months, 0.98 [0.03] at 36-42 months, and 1.00 [0.02] at 42-48 months) and can be used to support clinical and public health efforts focused on obesity prevention in early life., (© 2023. The Author(s).)

Infrared thermography is a non-destructive technique that can be exploited in many fields including polymer composite investigation. Based on emissivity and thermal diffusivity variation; components, defects, and curing state of the composite can be identified. However, manual processing of thermal images that may contain significant artifacts, is prone to erroneous component and property determination. In this study, thermal images of different graphite/graphene-based polymer composites fabricated by hand, planetary, and batch mixing techniques were analyzed through an automatic machine learning model. Filler size, shape, and location can be identified in polymer composites and thus, the dispersion of different samples was quantified with a resolution of ~ 20 µm despite having artifacts in the thermal image. Thermal diffusivity comparison of three mixing techniques was performed for 40% graphite in the elastomer. Batch mixing demonstrated superior dispersion than planetary and hand mixing as the dispersion index (DI) for batch mixing was 0.07 while planetary and hand mixing showed 0.0865 and 0.163 respectively. Curing was investigated for a polymer with different fillers (PDMS took 500 s while PDMS-Graphene and PDMS Graphite Powder took 800 s to cure), and a thermal characteristic curve was generated to compare the composite quality. Therefore, the above-mentioned methods with machine learning algorithms can be a great tool to analyze composite both quantitatively and qualitatively., (© 2023. The Author(s).)

The emergence of genome-wide association studies (GWAS) has led to the creation of large repositories of human genetic variation, creating enormous opportunities for genetic research and worldwide collaboration. Methods that are based on GWAS summary statistics seek to leverage such records, overcoming barriers that often exist in individual-level data access while also offering significant computational savings. Such summary-statistics-based applications include GWAS meta-analysis, with and without sample overlap, and case-case GWAS. We compare performance of leading methods for summary-statistics-based genomic analysis and also introduce a novel framework that can unify usual summary-statistics-based implementations via the reconstruction of allelic and genotypic frequencies and counts (ReACt). First, we evaluate ASSET, METAL, and ReACt using both synthetic and real data for GWAS meta-analysis (with and without sample overlap) and find that, while all three methods are comparable in terms of power and error control, ReACt and METAL are faster than ASSET by a factor of at least hundred. We then proceed to evaluate performance of ReACt vs an existing method for case-case GWAS and show comparable performance, with ReACt requiring minimal underlying assumptions and being more user-friendly. Finally, ReACt allows us to evaluate, for the first time, an implementation for calculating polygenic risk score (PRS) for groups of cases and controls based on summary statistics. Our work demonstrates the power of GWAS summary-statistics-based methodologies and the proposed novel method provides a unifying framework and allows further extension of possibilities for researchers seeking to understand the genetics of complex disease., (© 2022. The Author(s).)

Axonal characterizations of connectomes in healthy and disease phenotypes are surprisingly incomplete and biased because unmyelinated axons, the most prevalent type of fibers in the nervous system, have largely been ignored as their quantitative assessment quickly becomes unmanageable as the number of axons increases. Herein, we introduce the first prototype of a high-throughput processing pipeline for automated segmentation of unmyelinated fibers. Our team has used transmission electron microscopy images of vagus and pelvic nerves in rats. All unmyelinated axons in these images are individually annotated and used as labeled data to train and validate a deep instance segmentation network. We investigate the effect of different training strategies on the overall segmentation accuracy of the network. We extensively validate the segmentation algorithm as a stand-alone segmentation tool as well as in an expert-in-the-loop hybrid segmentation setting with preliminary, albeit remarkably encouraging results. Our algorithm achieves an instance-level [Formula: see text] score of between 0.7 and 0.9 on various test images in the stand-alone mode and reduces expert annotation labor by 80% in the hybrid setting. We hope that this new high-throughput segmentation pipeline will enable quick and accurate characterization of unmyelinated fibers at scale and become instrumental in significantly advancing our understanding of connectomes in both the peripheral and the central nervous systems., (© 2022. The Author(s).)

Protein 3D structure prediction has advanced significantly in recent years due to improving contact prediction accuracy. This improvement has been largely due to deep learning approaches that predict inter-residue contacts and, more recently, distances using multiple sequence alignments (MSAs). In this work we present AttentiveDist, a novel approach that uses different MSAs generated with different E-values in a single model to increase the co-evolutionary information provided to the model. To determine the importance of each MSA's feature at the inter-residue level, we added an attention layer to the deep neural network. We show that combining four MSAs of different E-value cutoffs improved the model prediction performance as compared to single E-value MSA features. A further improvement was observed when an attention layer was used and even more when additional prediction tasks of bond angle predictions were added. The improvement of distance predictions were successfully transferred to achieve better protein tertiary structure modeling.

Potential inhibitors of a target biomolecule, NAD-dependent deacetylase Sirtuin 1, were identified by a contest-based approach, in which participants were asked to propose a prioritized list of 400 compounds from a designated compound library containing 2.5 million compounds using in silico methods and scoring. Our aim was to identify target enzyme inhibitors and to benchmark computer-aided drug discovery methods under the same experimental conditions. Collecting compound lists derived from various methods is advantageous for aggregating compounds with structurally diversified properties compared with the use of a single method. The inhibitory action on Sirtuin 1 of approximately half of the proposed compounds was experimentally accessed. Ultimately, seven structurally diverse compounds were identified.

With advancements in synthetic biology, the cost and the time needed for designing and synthesizing customized gene products have been steadily decreasing. Many research laboratories in academia as well as industry routinely create genetically engineered proteins as a part of their research activities. However, manipulation of protein sequences could result in unintentional production of toxic proteins. Therefore, being able to identify the toxicity of a protein before the synthesis would reduce the risk of potential hazards. Existing methods are too specific, which limits their application. Here, we extended general function prediction methods for predicting the toxicity of proteins. Protein function prediction methods have been actively studied in the bioinformatics community and have shown significant improvement over the last decade. We have previously developed successful function prediction methods, which were shown to be among top-performing methods in the community-wide functional annotation experiment, CAFA. Based on our function prediction method, we developed a neural network model, named NNTox, which uses predicted GO terms for a target protein to further predict the possibility of the protein being toxic. We have also developed a multi-label model, which can predict the specific toxicity type of the query sequence. Together, this work analyses the relationship between GO terms and protein toxicity and builds predictor models of protein toxicity.

Protein-protein interactions (PPIs) play essential roles in many biological processes. A PPI network provides crucial information on how biological pathways are structured and coordinated from individual protein functions. In the past two decades, large-scale PPI networks of a handful of organisms were determined by experimental techniques. However, these experimental methods are time-consuming, expensive, and are not easy to perform on new target organisms. Large-scale PPI data is particularly sparse in plant organisms. Here, we developed a computational approach for detecting PPIs trained and tested on known PPIs of Arabidopsis thaliana and applied to three plants, Arabidopsis thaliana, Glycine max (soybean), and Zea mays (maize) to discover new PPIs on a genome-scale. Our method considers a variety of features including protein sequences, gene co-expression, functional association, and phylogenetic profiles. This is the first work where a PPI prediction method was developed for is the first PPI prediction method applied on benchmark datasets of Arabidopsis. The method showed a high prediction accuracy of over 90% and very high precision of close to 1.0. We predicted 50,220 PPIs in Arabidopsis thaliana, 13,175,414 PPIs in corn, and 13,527,834 PPIs in soybean. Newly predicted PPIs were classified into three confidence levels according to the availability of existing supporting evidence and discussed. Predicted PPIs in the three plant genomes are made available for future reference.

The problem of reverse-engineering the evolution of a dynamic network, known broadly as network archaeology, is one of profound importance in diverse application domains. In analysis of infection spread, it reveals the spatial and temporal processes underlying infection. In analysis of biomolecular interaction networks (e.g., protein interaction networks), it reveals early molecules that are known to be differentially implicated in diseases. In economic networks, it reveals flow of capital and associated actors. Beyond these recognized applications, it provides analytical substrates for novel studies - for instance, on the structural and functional evolution of the human brain connectome. In this paper, we model, formulate, and rigorously analyze the problem of inferring the arrival order of nodes in a dynamic network from a single snapshot. We derive limits on solutions to the problem, present methods that approach this limit, and demonstrate the methods on a range of applications, from inferring the evolution of the human brain connectome to conventional citation and social networks, where ground truth is known.

We have developed a framework to convert an arbitrary integer factorization problem to an executable Ising model by first writing it as an optimization function then transforming the k-bit coupling (k ≥ 3) terms to quadratic terms using ancillary variables. Our resource-efficient method uses [Formula: see text] binary variables (qubits) for finding the factors of an integer N. We present how to factorize 15, 143, 59989, and 376289 using 4, 12, 59, and 94 logical qubits, respectively. This method was tested using the D-Wave 2000Q for finding an embedding and determining the prime factors for a given composite number. The method is general and could be used to factor larger integers as the number of available qubits increases, or combined with other ad hoc methods to achieve better performances for specific numbers.

The study of network topology provides insight into the function and behavior of physical, social, and biological systems. A natural step towards discovering the organizing principles of these complex topologies is to identify a reduced network representation using cohesive subgroups or communities. This procedure often uncovers the underlying mechanisms governing the functional assembly of complex networks. A community is usually defined as a subgraph or a set of nodes that has more edges than would be expected from a simple, null distribution of edges over the graph. This view drives objective such as modularity. Another perspective, corresponding to objectives like conductance or density, is that communities are groups of nodes that have extremal properties with respect to the number of internal edges and cut edges. Here we show that identifying community boundaries rather than communities results in a more accurate decomposition of the network into informative components. We derive a network analog of Gauss's law that relates a measure of flux through a subgraph's boundary to the connectivity among the subgraph's nodes. Our Gauss's law for networks naturally characterizes a community as a subgraph with high flux through its boundary. Aggregating flux over these boundaries gives rise to a Laplacian and forms the basis of our "Laplacian modularity" quality function for community detection that is applicable to general network types. This technique allows us to determine communities that are both overlapping and hierarchically organized.

We propose a new iterative screening contest method to identify target protein inhibitors. After conducting a compound screening contest in 2014, we report results acquired from a contest held in 2015 in this study. Our aims were to identify target enzyme inhibitors and to benchmark a variety of computer-aided drug discovery methods under identical experimental conditions. In both contests, we employed the tyrosine-protein kinase Yes as an example target protein. Participating groups virtually screened possible inhibitors from a library containing 2.4 million compounds. Compounds were ranked based on functional scores obtained using their respective methods, and the top 181 compounds from each group were selected. Our results from the 2015 contest show an improved hit rate when compared to results from the 2014 contest. In addition, we have successfully identified a statistically-warranted method for identifying target inhibitors. Quantitative analysis of the most successful method gave additional insights into important characteristics of the method used.

Protein tertiary structure prediction methods have matured in recent years. However, some proteins defy accurate prediction due to factors such as inadequate template structures. While existing model quality assessment methods predict global model quality relatively well, there is substantial room for improvement in local quality assessment, i.e. assessment of the error at each residue position in a model. Local quality is a very important information for practical applications of structure models such as interpreting/designing site-directed mutagenesis of proteins. We have developed a novel local quality assessment method for protein tertiary structure models. The method, named Graph-based Model Quality assessment method (GMQ), explicitly considers the predicted quality of spatially neighboring residues using a graph representation of a query protein structure model. GMQ uses conditional random field as its core of the algorithm, and performs a binary prediction of the quality of each residue in a model, indicating if a residue position is likely to be within an error cutoff or not. The accuracy of GMQ was improved by considering larger graphs to include quality information of more surrounding residues. Moreover, we found that using different edge weights in graphs reflecting different secondary structures further improves the accuracy. GMQ showed competitive performance on a benchmark for quality assessment of structure models from the Critical Assessment of Techniques for Protein Structure Prediction (CASP).

We present EP-DNN, a protocol for predicting enhancers based on chromatin features, in different cell types. Specifically, we use a deep neural network (DNN)-based architecture to extract enhancer signatures in a representative human embryonic stem cell type (H1) and a differentiated lung cell type (IMR90). We train EP-DNN using p300 binding sites, as enhancers, and TSS and random non-DHS sites, as non-enhancers. We perform same-cell and cross-cell predictions to quantify the validation rate and compare against two state-of-the-art methods, DEEP-ENCODE and RFECS. We find that EP-DNN has superior accuracy with a validation rate of 91.6%, relative to 85.3% for DEEP-ENCODE and 85.5% for RFECS, for a given number of enhancer predictions and also scales better for a larger number of enhancer predictions. Moreover, our H1 → IMR90 predictions turn out to be more accurate than IMR90 → IMR90, potentially because H1 exhibits a richer signature set and our EP-DNN model is expressive enough to extract these subtleties. Our work shows how to leverage the full expressivity of deep learning models, using multiple hidden layers, while avoiding overfitting on the training data. We also lay the foundation for exploration of cross-cell enhancer predictions, potentially reducing the need for expensive experimentation.

Here we consider using quantum annealing to solve Set Cover with Pairs (SCP), an NP-hard combinatorial optimization problem that plays an important role in networking, computational biology, and biochemistry. We show an explicit construction of Ising Hamiltonians whose ground states encode the solution of SCP instances. We numerically simulate the time-dependent Schrödinger equation in order to test the performance of quantum annealing for random instances and compare with that of simulated annealing. We also discuss explicit embedding strategies for realizing our Hamiltonian construction on the D-wave type restricted Ising Hamiltonian based on Chimera graphs. Our embedding on the Chimera graph preserves the structure of the original SCP instance and in particular, the embedding for general complete bipartite graphs and logical disjunctions may be of broader use than that the specific problem we deal with.

Reconstructing metabolic and signaling pathways is an effective way of interpreting a genome sequence. A challenge in a pathway reconstruction is that often genes in a pathway cannot be easily found, reflecting current imperfect information of the target organism. In this work, we developed a new method for finding missing genes, which integrates multiple features, including gene expression, phylogenetic profile, and function association scores. Particularly, for considering function association between candidate genes and neighboring proteins to the target missing gene in the network, we used Co-occurrence Association Score (CAS) and PubMed Association Score (PAS), which are designed for capturing functional coherence of proteins. We showed that adding CAS and PAS substantially improve the accuracy of identifying missing genes in the yeast enzyme-enzyme network compared to the cases when only the conventional features, gene expression, phylogenetic profile, were used. Finally, it was also demonstrated that the accuracy improves by considering indirect neighbors to the target enzyme position in the network using a proper network-topology-based weighting scheme.

Distributions of protein families and folds in genomes are highly skewed, having a small number of prevalent superfamiles/superfolds and a large number of families/folds of a small size. Why are the distributions of protein families and folds skewed? Why are there only a limited number of protein families? Here, we employ an information theoretic approach to investigate the protein sequence-structure relationship that leads to the skewed distributions. We consider that protein sequences and folds constitute an information theoretic channel and computed the most efficient distribution of sequences that code all protein folds. The identified distributions of sequences and folds are found to follow a power law, consistent with those observed for proteins in nature. Importantly, the skewed distributions of sequences and folds are suggested to have different origins: the skewed distribution of sequences is due to evolutionary pressure to achieve efficient coding of necessary folds, whereas that of folds is based on the thermodynamic stability of folds. The current study provides a new information theoretic framework for proteins that could be widely applied for understanding protein sequences, structures, functions, and interactions.