Application of transfer learning to predict drug-induced human in vivo gene expression changes using rat in vitro and in vivo data.

The liver is the primary site for the metabolism and detoxification of many compounds, including pharmaceuticals. Consequently, it is also the primary location for many adverse reactions. As the liver is not readily accessible for sampling in humans; rodent or cell line models are often used to evaluate potential toxic effects of a novel compound or candidate drug. However, relating the results of animal and in vitro studies to relevant clinical outcomes for the human in vivo situation still proves challenging. In this study, we incorporate principles of transfer learning within a deep artificial neural network allowing us to leverage the relative abundance of rat in vitro and in vivo exposure data from the Open TG-GATEs data set to train a model to predict the expected pattern of human in vivo gene expression following an exposure given measured human in vitro gene expression. We show that domain adaptation has been successfully achieved, with the rat and human in vitro data no longer being separable in the common latent space generated by the network. The network produces physiologically plausible predictions of human in vivo gene expression pattern following an exposure to a previously unseen compound. Moreover, we show the integration of the human in vitro data in the training of the domain adaptation network significantly improves the temporal accuracy of the predicted rat in vivo gene expression pattern following an exposure to a previously unseen compound. In this way, we demonstrate the improvements in prediction accuracy that can be achieved by combining data from distinct domains.

Blood Milieu in Acute Myocardial Infarction Reprograms Human Macrophages for Trauma Repair.

Acute myocardial infarction (AMI) is accompanied by a systemic trauma response that impacts the whole body, including blood. This study addresses whether macrophages, key players in trauma repair, sense and respond to these changes. For this, healthy human monocyte-derived macrophages are exposed to 20% human AMI (n = 50) or control (n = 20) serum and analyzed by transcriptional and multiparameter functional screening followed by network-guided data interpretation and drug repurposing. Results are validated in an independent cohort at functional level (n = 47 AMI, n = 25 control) and in a public dataset. AMI serum exposure results in an overt AMI signature, enriched in debris cleaning, mitosis, and immune pathways. Moreover, gene networks associated with AMI and with poor clinical prognosis in AMI are identified. Network-guided drug screening on the latter unveils prostaglandin E2 (PGE2) signaling as target for clinical intervention in detrimental macrophage imprinting during AMI trauma healing. The results demonstrate pronounced context-induced macrophage reprogramming by the AMI systemic environment, to a degree decisive for patient prognosis. This offers new opportunities for targeted intervention and optimized cardiovascular disease risk management.

Integrating multiplex immunofluorescent and mass spectrometry imaging to map myeloid heterogeneity in its metabolic and cellular context.

Cells often adopt different phenotypes, dictated by tissue-specific or local signals such as cell-cell and cell-matrix contacts or molecular micro-environment. This holds in extremis for macrophages with their high phenotypic plasticity. Their broad range of functions, some even opposing, reflects their heterogeneity, and a multitude of subsets has been described in different tissues and diseases. Such micro-environmental imprint cannot be adequately studied by single-cell applications, as cells are detached from their context, while histology-based assessment lacks the phenotypic depth due to limitations in marker combination. Here, we present a novel, integrative approach in which 15-color multispectral imaging allows comprehensive cell classification based on multi-marker expression patterns, followed by downstream analysis pipelines to link their phenotypes to contextual, micro-environmental cues, such as their cellular ("community") and metabolic ("local lipidome") niches in complex tissue. The power of this approach is illustrated for myeloid subsets and associated lipid signatures in murine atherosclerotic plaque.

Integrative multiomics analysis of human atherosclerosis reveals a serum response factor-driven network associated with intraplaque hemorrhage.

BACKGROUND: While single-omics analyses on human atherosclerotic plaque have been very useful to map stage- or disease-related differences in expression, they only partly capture the array of changes in this tissue and suffer from scale-intrinsic limitations. In order to better identify processes associated with intraplaque hemorrhage and plaque instability, we therefore combined multiple omics into an integrated model. METHODS: In this study, we compared protein and gene makeup of low- versus high-risk atherosclerotic lesion segments from carotid endarterectomy patients, as judged from the absence or presence of intraplaque hemorrhage, respectively. Transcriptomic, proteomic, and peptidomic data of this plaque cohort were aggregated and analyzed by DIABLO, an integrative multivariate classification and feature selection method. RESULTS: We identified a protein-gene associated multiomics model able to segregate stable, nonhemorrhaged from vulnerable, hemorrhaged lesions at high predictive performance (AUC >0.95). The dominant component of this model correlated with αSMA- PDGFRα+ fibroblast-like cell content (p = 2.4E-05) and Arg1+ macrophage content (p = 2.2E-04) and was driven by serum response factor (SRF), possibly in a megakaryoblastic leukemia-1/2 (MKL1/2) dependent manner. Gene set overrepresentation analysis on the selected key features of this model pointed to a clear cardiovascular disease signature, with overrepresentation of extracellular matrix synthesis and organization, focal adhesion, and cholesterol metabolism terms, suggestive of the model's relevance for the plaque vulnerability. Finally, we were able to corroborate the predictive power of the selected features in several independent mRNA and proteomic plaque cohorts. CONCLUSIONS: In conclusion, our integrative omics study has identified an intraplaque hemorrhage-associated cardiovascular signature that provides excellent stratification of low- from high-risk carotid artery plaques in several independent cohorts. Further study revealed suppression of an SRF-regulated disease network, controlling lesion stability, in vulnerable plaque, which can serve as a scaffold for the design of targeted intervention in plaque destabilization.

Use of deep learning methods to translate drug-induced gene expression changes from rat to human primary hepatocytes.

In clinical trials, animal and cell line models are often used to evaluate the potential toxic effects of a novel compound or candidate drug before progressing to human trials. However, relating the results of animal and in vitro model exposures to relevant clinical outcomes in the human in vivo system still proves challenging, relying on often putative orthologs. In recent years, multiple studies have demonstrated that the repeated dose rodent bioassay, the current gold standard in the field, lacks sufficient sensitivity and specificity in predicting toxic effects of pharmaceuticals in humans. In this study, we evaluate the potential of deep learning techniques to translate the pattern of gene expression measured following an exposure in rodents to humans, circumventing the current reliance on orthologs, and also from in vitro to in vivo experimental designs. Of the applied deep learning architectures applied in this study the convolutional neural network (CNN) and a deep artificial neural network with bottleneck architecture significantly outperform classical machine learning techniques in predicting the time series of gene expression in primary human hepatocytes given a measured time series of gene expression from primary rat hepatocytes following exposure in vitro to a previously unseen compound across multiple toxicologically relevant gene sets. With a reduction in average mean absolute error across 76 genes that have been shown to be predictive for identifying carcinogenicity from 0.0172 for a random regression forest to 0.0166 for the CNN model (p < 0.05). These deep learning architecture also perform well when applied to predict time series of in vivo gene expression given measured time series of in vitro gene expression for rats.

Heterogeneous Domain Adaptation for IHC Classification of Breast Cancer Subtypes.

Increasingly, multiple parallel omics datasets are collected from biological samples. Integrating these datasets for classification is an open area of research. Additionally, whilst multiple datasets may be available for the training samples, future samples may only be measured by a single technology requiring methods which do not rely on the presence of all datasets for sample prediction. This enables us to directly compare the protein and the gene profiles. New samples with just one set of measurements (e.g., just protein) can then be mapped to this latent common space where classification is performed. Using this approach, we achieved an improvement of up to 12 percent in accuracy when classifying samples based on their protein measurements compared with baseline methods which were trained on the protein data alone. We illustrate that the additional inclusion of the gene expression or protein expression in the training process enabled the separation between the classes to become clearer.

Novel concept to guide systolic heart failure medication by repeated biomarker testing-results from TIME-CHF in context of predictive, preventive, and personalized medicine.

BACKGROUND: It is uncertain whether repeated measurements of a multi-target biomarker panel may help to personalize medical heart failure (HF) therapy to improve outcome in chronic HF. METHODS: This analysis included 499 patients from the Trial of Intensified versus standard Medical therapy in Elderly patients with Congestive Heart Failure (TIME-CHF), aged ≥ 60 years, LVEF ≤ 45%, and NYHA ≥ II, who had repeated clinical visits within 19 months follow-up. The interaction between repeated measurements of biomarkers and treatment effects of loop diuretics, spironolactone, ß-blockers, and renin-angiotensin system (RAS) inhibitors on risk of HF hospitalization or death was investigated in a hypothesis-generating analysis. Generalized estimating equation (GEE) models were used to account for the correlation between recurrences of events in a patient. RESULTS: One hundred patients (20%) had just one event (HF hospitalization or death) and 87 (17.4%) had at least two events. Loop diuretic up-titration had a beneficial effect for patients with high interleukin-6 (IL6) or high high-sensitivity C-reactive protein (hsCRP) (interaction, P = 0.013 and P = 0.001), whereas the opposite was the case with low hsCRP (interaction, P = 0.013). Higher dosage of loop diuretics was associated with poor outcome in patients with high blood urea nitrogen (BUN) or prealbumin (interaction, P = 0.006 and P = 0.001), but not in those with low levels of these biomarkers. Spironolactone up-titration was associated with lower risk of HF hospitalization or death in patients with high cystatin C (CysC) (interaction, P = 0.021). ß-Blockers up-titration might have a beneficial effect in patients with low soluble fms-like tyrosine kinase-1 (sFlt) (interaction, P = 0.021). No treatment biomarker interactions were found for RAS inhibition. CONCLUSION: The data of this post hoc analysis suggest that decision-making using repeated biomarker measurements may be very promising in bringing treatment of heart failure to a new level in the context of predictive, preventive, and personalized medicine. Clearly, prospective testing is needed before this novel concept can be adopted. CLINICAL TRIAL REGISTRATION: isrctn.org, identifier: ISRCTN43596477.