Search

Network biology is useful for modeling complex biological phenomena; it has attracted attention with the advent of novel graph-based machine learning methods. However, biological applications of network methods often suffer from inadequate follow-up. In this perspective, we discuss obstacles for contemporary network approaches—particularly focusing on challenges representing biological concepts, applying machine learning methods, and interpreting and validating computational findings about biology—in an effort to catalyze actionable biological discovery. [ABSTRACT FROM AUTHOR]

ABSTRACT: A novel computational model of implant migration in trabecular bone was developed using smoothed‐particle hydrodynamics (SPH), and an initial validation was performed via correlation with experimental data. Six fresh‐frozen human cadaveric specimens measuring 10 × 10 × 20 mm were extracted from the proximal femurs of female donors (mean age of 82 years, range 75–90, BV/TV ratios between 17.88% and 30.49%). These specimens were then penetrated under axial loading to a depth of 10 mm with 5 mm diameter cylindrical indenters bearing either flat or sharp/conical tip designs similar to blunt and self‐tapping cancellous screws, assigned in a random manner. SPH models were constructed based on microCT scans (17.33 µm) of the cadaveric specimens. Two initial specimens were used for calibration of material model parameters. The remaining four specimens were then simulated in silico using identical material model parameters. Peak forces varied between 92.0 and 365.0 N in the experiments, and 115.5–352.2 N in the SPH simulations. The concordance correlation coefficient between experimental and simulated pairs was 0.888, with a 95%CI of 0.8832–0.8926, a Pearson ρ (precision) value of 0.9396, and a bias correction factor Cb (accuracy) value of 0.945. Patterns of bone compaction were qualitatively similar; both experimental and simulated flat‐tipped indenters produced dense regions of compacted material adjacent to the advancing face of the indenter, while sharp‐tipped indenters deposited compacted material along their peripheries. Simulations based on SPH can produce accurate predictions of trabecular bone penetration that are useful for characterizing implant performance under high‐strain loading conditions. © 2017 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 36:1114–1123, 2018. [ABSTRACT FROM AUTHOR]

Identifying noncoding single nucleotide variants ( SNVs ) in regulatory DNA linked to polygenic disease risk, the transcription factors ( TFs ) they bind, and the target genes they dysregulate is a goal in polygenic disease research. Massively parallel reporter gene analysis ( MPRA ) of 3,451 SNVs linked to risk for polygenic skin diseases characterized by disrupted epidermal homeostasis identified 355 differentially active SNVs ( daSNVs ). daSNV target gene analysis, combined with daSNV editing, underscored dysregulated epidermal differentiation as a pathomechanism shared across common polygenic skin diseases. CRISPR knockout screens of 1772 human TFs revealed 108 TFs essential for epidermal progenitor differentiation, uncovering novel roles for ZNF217, CXXC1, FOXJ2, IRX2 and NRF1. Population sampling CUT&RUN of 27 homeostatic TFs identified allele-specific DNA binding ( ASB ) differences at daSNVs enriched near epidermal homeostasis and monogenic skin disease genes, with notable representation of SP/KLF and AP-1/2 TFs. This resource implicates dysregulated differentiation in risk for diverse polygenic skin diseases.

Viral proteins localize within subcellular compartments to subvert host machinery and promote pathogenesis. To study SARS-CoV-2 biology, we generated an atlas of 2422 human proteins vicinal to 17 SARS-CoV-2 viral proteins using proximity proteomics. This identified viral proteins at specific intracellular locations, such as association of accessary proteins with intracellular membranes, and projected SARS-CoV-2 impacts on innate immune signaling, ER-Golgi transport, and protein translation. It identified viral protein adjacency to specific host proteins whose regulatory variants are linked to COVID-19 severity, including the TRIM4 interferon signaling regulator which was found proximal to the SARS-CoV-2 M protein. Viral NSP1 protein adjacency to the EIF3 complex was associated with inhibited host protein translation whereas ORF6 localization with MAVS was associated with inhibited RIG-I 2CARD-mediated IFNB1 promoter activation. Quantitative proteomics identified candidate host targets for the NSP5 protease, with specific functional cleavage sequences in host proteins CWC22 and FANCD2. This data resource identifies host factors proximal to viral proteins in living human cells and nominates pathogenic mechanisms employed by SARS-CoV-2., Author Summary: SARS-CoV-2 is the latest pathogenic coronavirus to emerge as a public health threat. We create a database of proximal host proteins to 17 SARS-CoV-2 viral proteins. We validate that NSP1 is proximal to the EIF3 translation initiation complex and is a potent inhibitor of translation. We also identify ORF6 antagonism of RNA-mediate innate immune signaling. We produce a database of potential host targets of the viral protease NSP5, and create a fluorescence-based assay to screen cleavage of peptide sequences. We believe that this data will be useful for identifying roles for many of the uncharacterized SARS-CoV-2 proteins and provide insights into the pathogenicity of new or emerging coronaviruses.

One in five Americans experience mental illness, and roughly 75% of psychiatric prescriptions do not successfully treat the patient's condition. Extensive evidence implicates genetic factors and signaling disruption in the pathophysiology of these diseases. Changes in transcription often underlie this molecular pathway dysregulation; individual patient transcriptional data can improve the efficacy of diagnosis and treatment. Recent large-scale genomic studies have uncovered shared genetic modules across multiple psychiatric disorders - providing an opportunity for an integrated multi-disease approach for diagnosis. Moreover, network-based models informed by gene expression can represent pathological biological mechanisms and suggest new genes for diagnosis and treatment. Here, we use patient gene expression data from multiple studies to classify psychiatric diseases, integrate knowledge from expert-curated databases and publicly available experimental data to create augmented disease-specific gene sets, and use these to recommend disease-relevant drugs. From Gene Expression Omnibus, we extract expression data from 145 cases of schizophrenia, 82 cases of bipolar disorder, 190 cases of major depressive disorder, and 307 shared controls. We use pathway-based approaches to predict psychiatric disease diagnosis with a random forest model (78% accuracy) and derive important features to augment available drug and disease signatures. Using protein-protein-interaction networks and embedding-based methods, we build a pipeline to prioritize treatments for psychiatric diseases that achieves a 3.4-fold improvement over a background model. Thus, we demonstrate that gene-expression-derived pathway features can diagnose psychiatric diseases and that molecular insights derived from this classification task can inform treatment prioritization for psychiatric diseases.

Millions of Americans are affected by rare diseases, many of which have poor survival rates. However, the small market size of individual rare diseases, combined with the time and capital requirements of pharmaceutical R&D, have hindered the development of new drugs for these cases. A promising alternative is drug repurposing, whereby existing FDA-approved drugs might be used to treat diseases different from their original indications. In order to generate drug repurposing hypotheses in a systematic and comprehensive fashion, it is essential to integrate information from across the literature of pharmacology, genetics, and pathology. To this end, we leverage a newly developed knowledge graph, the Global Network of Biomedical Relationships (GNBR). GNBR is a large, heterogeneous knowledge graph comprising drug, disease, and gene (or protein) entities linked by a small set of semantic themes derived from the abstracts of biomedical literature. We apply a knowledge graph embedding method that explicitly models the uncertainty associated with literature-derived relationships and uses link prediction to generate drug repurposing hypotheses. This approach achieves high performance on a gold-standard test set of known drug indications (AUROC = 0.89) and is capable of generating novel repurposing hypotheses, which we independently validate using external literature sources and protein interaction networks. Finally, we demonstrate the ability of our model to produce explanations of its predictions.

Diabetic retinopathy is a leading cause of blindness among working-age adults. Early detection of this condition is critical for good prognosis. In this paper, we demonstrate the use of convolutional neural networks (CNNs) on color fundus images for the recognition task of diabetic retinopathy staging. Our network models achieved test metric performance comparable to baseline literature results, with validation sensitivity of 95%. We additionally explored multinomial classification models, and demonstrate that errors primarily occur in the misclassification of mild disease as normal due to the CNNs inability to detect subtle disease features. We discovered that preprocessing with contrast limited adaptive histogram equalization and ensuring dataset fidelity by expert verification of class labels improves recognition of subtle features. Transfer learning on pretrained GoogLeNet and AlexNet models from ImageNet improved peak test set accuracies to 74.5%, 68.8%, and 57.2% on 2-ary, 3-ary, and 4-ary classification models, respectively.

Procedural sedation has allowed many painful interventions to be conducted outside the operating room. During such procedures, it is important to maintain an appropriate level of sedation to minimize the risk of respiratory depression if patients are over-sedated and added pain or anxiety if under-sedated. However, there is currently no objective way to measure the patient's evolving level of sedation during a procedure. We investigated the use of capnography-derived features as an objective measure of sedation level. Time-based capnograms were recorded from 30 patients during sedation for cardioversion. Through causal k-means clustering of selected features, we sequentially assigned each exhalation to one of three distinct clusters, or states. Transitions between these states correlated to events during sedation (drug administration, procedure start and end, and clinical interventions). Similar clustering of capnogram recordings from 26 healthy, non-sedated subjects did not reveal distinctly separated states.