Dataset Information

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

ABSTRACT: 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.

SUBMITTER: O'Donovan SD

PROVIDER: S-EPMC7418976 | biostudies-literature | 2020

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

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

O'Donovan Shauna D SD Driessens Kurt K Lopatta Daniel D Wimmenauer Florian F Lukas Alexander A Neeven Jelmer J Stumm Tobias T Smirnov Evgueni E Lenz Michael M Ertaylan Gokhan G Jennen Danyel G J DGJ van Riel Natal A W NAW Cavill Rachel R Peeters Ralf L M RLM de Kok Theo M C M TMCM

PloS one 20200811 8

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, l ...[more]

PMID: 32780735

Similar Datasets

Project description:AIM:To investigate the effects of electroporation on primary rat hepatocyte and to optimize the electroporation conditions introducing foreign genes into primary hepatocytes. METHODS:A single-pulse procedure was performed at low voltage (220-400 V) but with high capacitance (500-950 microF). Hepatocytes were divided into 4 groups according to the electroporation conditions: group I, 220 V and 500 microF; group II, 220 V and 950 microF; group III, 400 V and 950 microF,and group IV. The control group was freshly isolated hepatocytes and directly cultured under the same conditions as those of electroporation groups. The effects of electroporation on primary rat hepatocytes were detected by trypan blue exclusion (TBE) and MTT analysis. Besides, albumin (Alb), alanine transaminase (ALT) and lactate dehydrogenase (LDH) in the supernatants of cultured hepatocytes were measured by biochemical assay. RESULTS:Between day 1 and day 15 after incubation, primary rat hepatocytes of each electroporation group appeared normal, being the same with those of control group. TBE staining showed that slight hepatocyte damage and high survival rate were found in the electroporation groups and the control group. Cultured for 3, 7, 11 and 15 days, hepatocyte viability was approximately 92.6+/-2.5 %, 89.5+/-3.3 %, 82.0+/-3.5 % and 74.3+/-1.2 %, respectively. MTT analysis indicated that the viabilities of hepatocytes had no significant difference between each electroporation group, and those were similar to that of control group. At the 36th hour after electroporation, Alb, ALT and LDH in the supernatants of control group were 5.3+/-0.1 g x L(-1), 183.7+/-8.4 nkat x L(-1) and 896.8+/-58.5 nkat x L(-1); those of group II were 5.7+/-0.1 g x L(-1), 215.4+/-16.7 nkat x L(-1) and 1063.8+/-51.8 nkat x L(-1); and those of group III were 5.8+/-0.2 g x L(-1), 217.1+/-8.4 nkat x L(-1) and 1063.8+/-10.0 nkat x L(-1). Statistically, the proteins of group II and group III were significantly higher than those of control group (P<0.05), whereas the protein production of group I, Alb, ALT and LDH were 5.3+/-0.2 g x L(-1), 205.4+/-3.3 nkat x L(-1) and 1035.4+/-116.9 nkat x L(-1), were similar to those of control group. At the same time, TBE and MTT analysis indicated that there was no significant cell viability difference between electroporation groups and control group. CONCLUSION:This single-pulse electroporation procedure performed at low voltage (220-400 V) but with high capacitance (950 microF) is one of the optimal choices to introduce foreign genes into primary rat hepatocyte.

Project description:Machine learning methods have been applied to many data sets in pharmaceutical research for several decades. The relative ease and availability of fingerprint type molecular descriptors paired with Bayesian methods resulted in the widespread use of this approach for a diverse array of end points relevant to drug discovery. Deep learning is the latest machine learning algorithm attracting attention for many of pharmaceutical applications from docking to virtual screening. Deep learning is based on an artificial neural network with multiple hidden layers and has found considerable traction for many artificial intelligence applications. We have previously suggested the need for a comparison of different machine learning methods with deep learning across an array of varying data sets that is applicable to pharmaceutical research. End points relevant to pharmaceutical research include absorption, distribution, metabolism, excretion, and toxicity (ADME/Tox) properties, as well as activity against pathogens and drug discovery data sets. In this study, we have used data sets for solubility, probe-likeness, hERG, KCNQ1, bubonic plague, Chagas, tuberculosis, and malaria to compare different machine learning methods using FCFP6 fingerprints. These data sets represent whole cell screens, individual proteins, physicochemical properties as well as a data set with a complex end point. Our aim was to assess whether deep learning offered any improvement in testing when assessed using an array of metrics including AUC, F1 score, Cohen's kappa, Matthews correlation coefficient and others. Based on ranked normalized scores for the metrics or data sets Deep Neural Networks (DNN) ranked higher than SVM, which in turn was ranked higher than all the other machine learning methods. Visualizing these properties for training and test sets using radar type plots indicates when models are inferior or perhaps over trained. These results also suggest the need for assessing deep learning further using multiple metrics with much larger scale comparisons, prospective testing as well as assessment of different fingerprints and DNN architectures beyond those used.

Project description:Sandwich-cultured rat hepatocytes are used in drug discovery for pharmacological and toxicological assessment of drug candidates, yet their utility as a functional model for drug transporters has not been fully characterized. To evaluate the system as an in vitro model for drug transport, expression changes of hepatic transporters relative to whole liver and freshly isolated hepatocytes (day 0) were examined by real-time quantitative reverse transcription-polymerase chain reaction for 4 consecutive days of culture. No significant differences in transporter expression levels were observed between freshly isolated hepatocytes and whole liver. Two distinct mRNA profiles were detected over time showing 1) a more than 5-fold decline in levels of uptake transporters such as Na(+)-taurocholate cotransporting polypeptide (Ntcp), organic anion transporter (Oat) 2, organic anion-transporting polypeptide (Oatp) 1a1, Oatp1a4, and Oatp1b2 and 2) a greater than 5-fold increase of efflux transporters P-glycoprotein (P-gp), breast cancer resistance protein (Bcrp), and multidrug resistance-related proteins (Mrp) 1, 2, 3, and 4. In addition, protein levels and functional activities for selected transporters were also determined. Protein levels for Mrp2, Bcrp, P-gp, Ntcp, and Oatp1a4 corresponded to changes in mRNA. Functional activities of Oatps and Oct1 exhibited a 3- and 4-fold decrease on day 2 and day 4, respectively, relative to that on day 0, whereas a more than 10-fold reduction in Oat2 activity was observed. These results indicate that the cell culture conditions used herein did not provide an optimal environment for expression of all hepatic transporters. Significant time-dependent alterations in basal gene expression patterns of transporters were detected compared with those in liver or freshly isolated hepatocytes. Further work and new strategies are required to improve the validity of this model as an in vitro tool for in vivo drug transport or biliary clearance prediction.

Project description:One of the main obstacles to the successful treatment of cancer is the phenomenon of drug resistance. A common strategy to overcome resistance is the use of combination therapies. However, the space of possibilities is huge and efficient search strategies are required. Machine Learning (ML) can be a useful tool for the discovery of novel, clinically relevant anti-cancer drug combinations. In particular, deep learning (DL) has become a popular choice for modeling drug combination effects. Here, we set out to examine the impact of different methodological choices on the performance of multimodal DL-based drug synergy prediction methods, including the use of different input data types, preprocessing steps and model architectures. Focusing on the NCI ALMANAC dataset, we found that feature selection based on prior biological knowledge has a positive impact-limiting gene expression data to cancer or drug response-specific genes improved performance. Drug features appeared to be more predictive of drug response, with a 41% increase in coefficient of determination (R2) and 26% increase in Spearman correlation relative to a baseline model that used only cell line and drug identifiers. Molecular fingerprint-based drug representations performed slightly better than learned representations-ECFP4 fingerprints increased R2 by 5.3% and Spearman correlation by 2.8% w.r.t the best learned representations. In general, fully connected feature-encoding subnetworks outperformed other architectures. DL outperformed other ML methods by more than 35% (R2) and 14% (Spearman). Additionally, an ensemble combining the top DL and ML models improved performance by about 6.5% (R2) and 4% (Spearman). Using a state-of-the-art interpretability method, we showed that DL models can learn to associate drug and cell line features with drug response in a biologically meaningful way. The strategies explored in this study will help to improve the development of computational methods for the rational design of effective drug combinations for cancer therapy.

Dataset Information

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

Publications

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

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets