Predicting drug response of tumors from integrated genomic profiles by deep neural networks

Artificial Intelligence for Inflammatory Bowel Diseases (IBD); Accurately Predicting Adverse Outcomes Using Machine Learning

BACKGROUND

Inflammatory Bowel Diseases with its complexity and heterogeneity could benefit from the increased application of Artificial Intelligence in clinical management.

AIM

To accurately predict adverse outcomes in patients with IBD using advanced computational models in a nationally representative dataset for potential use in clinical practice.

METHODS

We built a training model cohort and validated our result in a separate cohort. We used LASSO and Ridge regressions, Support Vector Machines, Random Forests and Neural Networks to balance between complexity and interpretability and analyzed their relative performances and reported the strongest predictors to the respective models. The participants in our study were patients with IBD selected from The OptumLabs® Data Warehouse (OLDW), a longitudinal, real-world data asset with de-identified administrative claims and electronic health record (EHR) data.

RESULTS

We included 72,178 and 69,165 patients in the training and validation set, respectively. In total, 4.1% of patients in the validation set were hospitalized, 2.9% needed IBD-related surgeries, 17% used long-term steroids and 13% of patients were initiated with biological therapy. Of the AI models we tested, the Random Forest and LASSO resulted in high accuracies (AUCs 0.70-0.92). Our artificial neural network performed similarly well in most of the models (AUCs 0.61-0.90).

CONCLUSIONS

This study demonstrates feasibility of accurately predicting adverse outcomes using complex and novel AI models on large longitudinal data sets of patients with IBD. These models could be applied for risk stratification and implementation of preemptive measures to avoid adverse outcomes in a clinical setting.

Multiplex gene quantification as digital markers for extremely rapid evaluation of chemo-drug sensitivity

Current administrations for precision drug uses are limited in evaluation speed. Here, we propose the use of multiplex gene-based digital markers for the extremely rapid personalized prediction of individual sensitivity to cancer drugs. We first screen the transcriptional profiles by applying two to three gene filters and scoring genes by their impact on drug sensitivity and finalize the gene lists by K-nearest neighbors cross-validation. The digital markers are cancer type dependent, are composed of tens to hundreds of gene expressions, and are rapidly quantified by reverse transcription quantitative real-time PCR (qRT-PCR) within 1–3 h after tumor sampling. The area under the receiver operating characteristic curve reached 0.88 when testing the performance of digital markers on organoids derived from colorectal cancer patient tumors. The algorithm and corresponding graphic user interface were developed to demonstrate the promise of digital markers for extremely rapid drug recommendation.

•

Non-targeted multiplex genes are screened as digital markers for drug sensitivity

•

Transcription level cohort of 10s to 100s genes predicts drug sensitivity

•

Digital markers are quantified using qRT-PCR within 1–3 h

•

Digital markers guide extremely rapid chemo-drug uses after patient hospitalization

In clinical cancer medicine, many patients require immediate chemotherapy after hospitalization. Current administrations for precision drug uses are limited in evaluation speed, including genomic sequencing and tumor organoid evaluation. An extremely rapid evaluation protocol is in high demand to realize drug recommendation within a few hours after tumor sampling. In this work, we have proposed an approach for extremely rapid and personalized drug recommendation.

A Survey of Autoencoder Algorithms to Pave the Diagnosis of Rare Diseases

Rare diseases (RDs) concern a broad range of disorders and can result from various origins. For a long time, the scientific community was unaware of RDs. Impressive progress has already been made for certain RDs; however, due to the lack of sufficient knowledge, many patients are not diagnosed. Nowadays, the advances in high-throughput sequencing technologies such as whole genome sequencing, single-cell and others, have boosted the understanding of RDs. To extract biological meaning using the data generated by these methods, different analysis techniques have been proposed, including machine learning algorithms. These methods have recently proven to be valuable in the medical field. Among such approaches, unsupervised learning methods via neural networks including autoencoders (AEs) or variational autoencoders (VAEs) have shown promising performances with applications on various type of data and in different contexts, from cancer to healthy patient tissues. In this review, we discuss how AEs and VAEs have been used in biomedical settings. Specifically, we discuss their current applications and the improvements achieved in diagnostic and survival of patients. We focus on the applications in the field of RDs, and we discuss how the employment of AEs and VAEs would enhance RD understanding and diagnosis.

Representation of molecules for drug response prediction

The rapid development of machine learning and deep learning algorithms in the recent decade has spurred an outburst of their applications in many research fields. In the chemistry domain, machine learning has been widely used to aid in drug screening, drug toxicity prediction, quantitative structure-activity relationship prediction, anti-cancer synergy score prediction, etc. This review is dedicated to the application of machine learning in drug response prediction. Specifically, we focus on molecular representations, which is a crucial element to the success of drug response prediction and other chemistry-related prediction tasks. We introduce three types of commonly used molecular representation methods, together with their implementation and application examples. This review will serve as a brief introduction of the broad field of molecular representations.

Predicting chemotherapy response using a variational autoencoder approach

Background

Multiple 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.

Results

We 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.

Conclusions

Given 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.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-021-04339-6.

Predicting drug sensitivity of cancer cells based on DNA methylation levels

Cancer cell lines, which are cell cultures derived from tumor samples, represent one of the least expensive and most studied preclinical models for drug development. Accurately predicting drug responses for a given cell line based on molecular features may help to optimize drug-development pipelines and explain mechanisms behind treatment responses. In this study, we focus on DNA methylation profiles as one type of molecular feature that is known to drive tumorigenesis and modulate treatment responses. Using genome-wide, DNA methylation profiles from 987 cell lines in the Genomics of Drug Sensitivity in Cancer database, we used machine-learning algorithms to evaluate the potential to predict cytotoxic responses for eight anti-cancer drugs. We compared the performance of five classification algorithms and four regression algorithms representing diverse methodologies, including tree-, probability-, kernel-, ensemble-, and distance-based approaches. We artificially subsampled the data to varying degrees, aiming to understand whether training based on relatively extreme outcomes would yield improved performance. When using classification or regression algorithms to predict discrete or continuous responses, respectively, we consistently observed excellent predictive performance when the training and test sets consisted of cell-line data. Classification algorithms performed best when we trained the models using cell lines with relatively extreme drug-response values, attaining area-under-the-receiver-operating-characteristic-curve values as high as 0.97. The regression algorithms performed best when we trained the models using the full range of drug-response values, although this depended on the performance metrics we used. Finally, we used patient data from The Cancer Genome Atlas to evaluate the feasibility of classifying clinical responses for human tumors based on models derived from cell lines. Generally, the algorithms were unable to identify patterns that predicted patient responses reliably; however, predictions by the Random Forests algorithm were significantly correlated with Temozolomide responses for low-grade gliomas.

A cross-level information transmission network for hierarchical omics data integration and phenotype prediction from a new genotype

MOTIVATION

An unsolved fundamental problem in biology is to predict phenotypes from a new genotype under environmental perturbations. The emergence of multiple omics data provides new opportunities but imposes great challenges in the predictive modeling of genotype-phenotype associations. Firstly, the high-dimensionality of genomics data and the lack of coherent labeled data often make the existing supervised learning techniques less successful. Secondly, it is challenging to integrate heterogeneous omics data from different resources. Finally, few works have explicitly modeled the information transmission from DNA to phenotype, which involves multiple intermediate molecular types. Higher-level features (e.g. gene expression) usually have stronger discriminative and interpretable power than lower-level features (e.g. somatic mutation).

RESULTS

We propose a novel Cross-LEvel Information Transmission (CLEIT) network framework to address the above issues. CLEIT aims to represent the asymmetrical multi-level organization of the biological system by integrating multiple incoherent omics data and to improve the prediction power of low-level features. CLEIT first learns the latent representation of the high-level domain then uses it as ground-truth embedding to improve the representation learning of the low-level domain in the form of contrastive loss. Besides, CLEIT can leverage the unlabeled heterogeneous omics data to improve the generalizability of the predictive model. We demonstrate the effectiveness and significant performance boost of CLEIT in predicting anti-cancer drug sensitivity from somatic mutations via the assistance of gene expressions when compared with state-of-the-art methods. CLEIT provides a general framework to model information transmissions and integrate multi-modal data in a multi-level system.

AVAILABILITYAND IMPLEMENTATION

The source code is freely available at https://github.com/XieResearchGroup/CLEIT.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Interpretable deep recommender system model for prediction of kinase inhibitor efficacy across cancer cell lines

Computational models for drug sensitivity prediction have the potential to significantly improve personalized cancer medicine. Drug sensitivity assays, combined with profiling of cancer cell lines and drugs become increasingly available for training such models. Multiple methods were proposed for predicting drug sensitivity from cancer cell line features, some in a multi-task fashion. So far, no such model leveraged drug inhibition profiles. Importantly, multi-task models require a tailored approach to model interpretability. In this work, we develop DEERS, a neural network recommender system for kinase inhibitor sensitivity prediction. The model utilizes molecular features of the cancer cell lines and kinase inhibition profiles of the drugs. DEERS incorporates two autoencoders to project cell line and drug features into 10-dimensional hidden representations and a feed-forward neural network to combine them into response prediction. We propose a novel interpretability approach, which in addition to the set of modeled features considers also the genes and processes outside of this set. Our approach outperforms simpler matrix factorization models, achieving R [Formula: see text] 0.82 correlation between true and predicted response for the unseen cell lines. The interpretability analysis identifies 67 biological processes that drive the cell line sensitivity to particular compounds. Detailed case studies are shown for PHA-793887, XMD14-99 and Dabrafenib.

HiDRA: Hierarchical Network for Drug Response Prediction with Attention

Understanding differences in drug responses between patients is crucial for delivering effective cancer treatment. We describe an interpretable AI model for use in predicting drug responses in cancer cells at the gene, molecular pathway, and drug level, which we have called the hierarchical network for drug response prediction with attention. We found that the model shows better accuracy in predicting drugs having efficacy against a given cell line than other state-of-the-art methods, with a root mean squared error of 1.0064, a Pearson's correlation coefficient of 0.9307, and an R² value of 0.8647. We also confirmed that the model gives high attention to drug-target genes and cancer-related pathways when predicting a response. The validity of predicted results was proven by in vitro cytotoxicity assay. Overall, we propose that our hierarchical and interpretable AI-based model is capable of interpreting intrinsic characteristics of cancer cells and drugs for accurate prediction of cancer-drug responses.

Predicting and characterizing a cancer dependency map of tumors with deep learning

Genome-wide loss-of-function screens have revealed genes essential for cancer cell proliferation, called cancer dependencies. It remains challenging to link cancer dependencies to the molecular compositions of cancer cells or to unscreened cell lines and further to tumors. Here, we present DeepDEP, a deep learning model that predicts cancer dependencies using integrative genomic profiles. It uses a unique unsupervised pretraining that captures unlabeled tumor genomic representations to improve the learning of cancer dependencies. We demonstrated DeepDEP's improvement over conventional machine learning methods and validated the performance with three independent datasets. By systematic model interpretations, we extended the current dependency maps with functional characterizations of dependencies and a proof-of-concept in silico assay of synthetic essentiality. We applied DeepDEP to pan-cancer tumor genomics and built the first pan-cancer synthetic dependency map of 8000 tumors with clinical relevance. In summary, DeepDEP is a novel tool for investigating cancer dependency with rapidly growing genomic resources.

Machine learning predicts treatment sensitivity in multiple myeloma based on molecular and clinical information coupled with drug response

Providing treatment sensitivity stratification at the time of cancer diagnosis allows better allocation of patients to alternative treatment options. Despite many clinical and biological risk markers having been associated with variable survival in cancer, assessing the interplay of these markers through Machine Learning (ML) algorithms still remains to be fully explored. Here, we present a Multi Learning Training approach (MuLT) combining supervised, unsupervised and self-supervised learning algorithms, to examine the predictive value of heterogeneous treatment outcomes for Multiple Myeloma (MM). We show that gene expression values improve the treatment sensitivity prediction and recapitulates genetic abnormalities detected by Fluorescence in situ hybridization (FISH) testing. MuLT performance was assessed by cross-validation experiments, in which it predicted treatment sensitivity with 68.70% of AUC. Finally, simulations showed numerical evidences that in average 17.07% of patients could get better response to a different treatment at the first line.

Performance Comparisons of AlexNet and GoogLeNet in Cell Growth Inhibition IC50 Prediction

Drug responses in cancer are diverse due to heterogenous genomic profiles. Drug responsiveness prediction is important in clinical response to specific cancer treatments. Recently, multi-class drug responsiveness models based on deep learning (DL) models using molecular fingerprints and mutation statuses have emerged. However, for multi-class models for drug responsiveness prediction, comparisons between convolution neural network (CNN) models (e.g., AlexNet and GoogLeNet) have not been performed. Therefore, in this study, we compared the two CNN models, GoogLeNet and AlexNet, along with the least absolute shrinkage and selection operator (LASSO) model as a baseline model. We constructed the models by taking drug molecular fingerprints of drugs and cell line mutation statuses, as input, to predict high-, intermediate-, and low-class for half-maximal inhibitory concentration (IC50) values of the drugs in the cancer cell lines. Additionally, we compared the models in breast cancer patients as well as in an independent gastric cancer cell line drug responsiveness data. We measured the model performance based on the area under receiver operating characteristic (ROC) curves (AUROC) value. In this study, we compared CNN models for multi-class drug responsiveness prediction. The AlexNet and GoogLeNet showed better performances in comparison to LASSO. Thus, DL models will be useful tools for precision oncology in terms of drug responsiveness prediction.

Computational Probing the Methylation Sites Related to EGFR Inhibitor-Responsive Genes

The emergence of drug resistance is one of the main obstacles to the treatment of lung cancer patients with EGFR inhibitors. Here, to further understand the mechanism of EGFR inhibitors in lung cancer and offer novel therapeutic targets for anti-EGFR-inhibitor resistance via the deep mining of pharmacogenomics data, we associated DNA methylation with drug sensitivities for uncovering the methylation sites related to EGFR inhibitor sensitivity genes. Specifically, we first introduced a grouped regularized regression model (Group Least Absolute Shrinkage and Selection Operator, group lasso) to detect the genes that were closely related to EGFR inhibitor effectiveness. Then, we applied the classical regression model (lasso) to identify the methylation sites associated with the above drug sensitivity genes. The new model was validated on the well-known cancer genomics resource: CTRP. GeneHancer and Encyclopedia of DNA Elements (ENCODE) database searches indicated that the predicted methylation sites related to EGFR inhibitor sensitivity genes were related to regulatory elements. Moreover, the correlation analysis on sensitivity genes and predicted methylation sites suggested that the methylation sites located in the promoter region were more correlated with the expression of EGFR inhibitor sensitivity genes than those located in the enhancer region and the TFBS. Meanwhile, we performed differential expression analysis of genes and predicted methylation sites and found that changes in the methylation level of some sites may affect the expression of the corresponding EGFR inhibitor-responsive genes. Therefore, we supposed that the effectiveness of EGFR inhibitors in lung cancer may be improved by methylation modification in their sensitivity genes.

LncTx: A network-based method to repurpose drugs acting on the survival-related lncRNAs in lung cancer

Despite the fact that an increased amount of survival-related lncRNAs have been found in cancer, few drugs that target lncRNAs are approved for treatment. Here, we developed a network-based algorithm, LncTx, to repurpose the medications that potentially act on survival-related lncRNAs in lung cancer. We used eight survival-related lncRNAs derived from our previous study to test the efficacy of this method. LncTx calculates the shortest path length (proximity) between the drug targets and the lncRNA-correlated proteins in the protein-protein interaction network (interactome). LncTx contains seven different proximity measures, which are calculated in the unweighted or weighted interactome. First, to test the performance of LncTx in predicting correct indication of drugs, we benchmarked the proximity measures based on the accuracy of differentiating anticancer drugs from non-anticancer drugs. The closest proximity weighted by clustering coefficient (closestCC) has the best performance (AUC around 0.8) compared to other proximity measures across all survival-related lncRNAs. The majority of the other six proximity measures have decent performance as well, with AUC greater than 0.7. Second, to evaluate whether LncTx can repurpose the drugs effectively acting on the lncRNAs, we clustered the drugs according to their proximities by hierarchical clustering. The drugs with smaller proximity (proximal drugs) were proved to be more effective than the drugs with larger proximity (distal drugs). In conclusion, LncTx enables us to accurately identify anticancer drugs and can potentially be an index to repurpose effective agents acting on survival-related lncRNAs in lung cancer.

Machine learning in the prediction of cancer therapy

Resistance to therapy remains a major cause of cancer treatment failures, resulting in many cancer-related deaths. Resistance can occur at any time during the treatment, even at the beginning. The current treatment plan is dependent mainly on cancer subtypes and the presence of genetic mutations. Evidently, the presence of a genetic mutation does not always predict the therapeutic response and can vary for different cancer subtypes. Therefore, there is an unmet need for predictive models to match a cancer patient with a specific drug or drug combination. Recent advancements in predictive models using artificial intelligence have shown great promise in preclinical settings. However, despite massive improvements in computational power, building clinically useable models remains challenging due to a lack of clinically meaningful pharmacogenomic data. In this review, we provide an overview of recent advancements in therapeutic response prediction using machine learning, which is the most widely used branch of artificial intelligence. We describe the basics of machine learning algorithms, illustrate their use, and highlight the current challenges in therapy response prediction for clinical practice.

Predicting Anti-Cancer Drug Response by Finding Optimal Subset of Drugs

MOTIVATION

One of the most difficult challenges in precision medicine is determining the best treatment strategy for each patient based on personal information. Since drug response prediction in vitro is extremely expensive, time-consuming, and virtually impossible, and because there are so many cell lines and drug data, computational methods are needed.

RESULTS

MinDrug is a method for predicting anti-cancer drug response which try to identify the best subset of drugs that are the most similar to other drugs. MinDrug predicts the anti-cancer drug response on a new cell line using information from drugs in this subset and their connections to other drugs. MinDrug employs a heuristic star algorithm to identify an optimal subset of drugs and a regression technique known as Elastic-Net approaches to predict anti-cancer drug response in a new cell line. To test MinDrug, we use both statistical and biological methods to assess the selected drugs. MinDrug is also compared to four state-of-the-art approaches using various k-fold cross-validations on two large public datasets: GDSC and CCLE. MinDrug outperforms the other approaches in terms of precision, robustness, and speed. Furthermore, we compare the evaluation results of all the approaches with an external dataset with a statistical distribution that is not exactly the same as the training data. The results show that MinDrug continues to outperform the other approaches.

AVAILABILITY

MinDrug's source code can be found at https://github.com/yassaee/MinDrug.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Super.FELT: supervised feature extraction learning using triplet loss for drug response prediction with multi-omics data

BACKGROUND

Predicting the drug response of a patient is important for precision oncology. In recent studies, multi-omics data have been used to improve the prediction accuracy of drug response. Although multi-omics data are good resources for drug response prediction, the large dimension of data tends to hinder performance improvement. In this study, we aimed to develop a new method, which can effectively reduce the large dimension of data, based on the supervised deep learning model for predicting drug response.

RESULTS

We proposed a novel method called Supervised Feature Extraction Learning using Triplet loss (Super.FELT) for drug response prediction. Super.FELT consists of three stages, namely, feature selection, feature encoding using a supervised method, and binary classification of drug response (sensitive or resistant). We used multi-omics data including mutation, copy number aberration, and gene expression, and these were obtained from cell lines [Genomics of Drug Sensitivity in Cancer (GDSC), Cancer Cell Line Encyclopedia (CCLE), and Cancer Therapeutics Response Portal (CTRP)], patient-derived tumor xenografts (PDX), and The Cancer Genome Atlas (TCGA). GDSC was used for training and cross-validation tests, and CCLE, CTRP, PDX, and TCGA were used for external validation. We performed ablation studies for the three stages and verified that the use of multi-omics data guarantees better performance of drug response prediction. Our results verified that Super.FELT outperformed the other methods at external validation on PDX and TCGA and was good at cross-validation on GDSC and external validation on CCLE and CTRP. In addition, through our experiments, we confirmed that using multi-omics data is useful for external non-cell line data.

CONCLUSION

By separating the three stages, Super.FELT achieved better performance than the other methods. Through our results, we found that it is important to train encoders and a classifier independently, especially for external test on PDX and TCGA. Moreover, although gene expression is the most powerful data on cell line data, multi-omics promises better performance for external validation on non-cell line data than gene expression data. Source codes of Super.FELT are available at  https://github.com/DMCB-GIST/Super.FELT .

Current Advances and Limitations of Deep Learning in Anticancer Drug Sensitivity Prediction

Anticancer drug screening can accelerate drug discovery to save the lives of cancer patients, but cancer heterogeneity makes this screening challenging. The prediction of anticancer drug sensitivity is useful for anticancer drug development and the identification of biomarkers of drug sensitivity. Deep learning, as a branch of machine learning, is an important aspect of in silico research. Its outstanding computational performance means that it has been used for many biomedical purposes, such as medical image interpretation, biological sequence analysis, and drug discovery. Several studies have predicted anticancer drug sensitivity based on deep learning algorithms. The field of deep learning has made progress regarding model performance and multi-omics data integration. However, deep learning is limited by the number of studies performed and data sources available, so it is not perfect as a pre-clinical approach for use in the anticancer drug screening process. Improving the performance of deep learning models is a pressing issue for researchers. In this review, we introduce the research of anticancer drug sensitivity prediction and the use of deep learning in this research area. To provide a reference for future research, we also review some common data sources and machine learning methods. Lastly, we discuss the advantages and disadvantages of deep learning, as well as the limitations and future perspectives regarding this approach.

Learning curves for drug response prediction in cancer cell lines

BACKGROUND

Motivated by the size and availability of cell line drug sensitivity data, researchers have been developing machine learning (ML) models for predicting drug response to advance cancer treatment. As drug sensitivity studies continue generating drug response data, a common question is whether the generalization performance of existing prediction models can be further improved with more training data.

METHODS

We utilize empirical learning curves for evaluating and comparing the data scaling properties of two neural networks (NNs) and two gradient boosting decision tree (GBDT) models trained on four cell line drug screening datasets. The learning curves are accurately fitted to a power law model, providing a framework for assessing the data scaling behavior of these models.

RESULTS

The curves demonstrate that no single model dominates in terms of prediction performance across all datasets and training sizes, thus suggesting that the actual shape of these curves depends on the unique pair of an ML model and a dataset. The multi-input NN (mNN), in which gene expressions of cancer cells and molecular drug descriptors are input into separate subnetworks, outperforms a single-input NN (sNN), where the cell and drug features are concatenated for the input layer. In contrast, a GBDT with hyperparameter tuning exhibits superior performance as compared with both NNs at the lower range of training set sizes for two of the tested datasets, whereas the mNN consistently performs better at the higher range of training sizes. Moreover, the trajectory of the curves suggests that increasing the sample size is expected to further improve prediction scores of both NNs. These observations demonstrate the benefit of using learning curves to evaluate prediction models, providing a broader perspective on the overall data scaling characteristics.

CONCLUSIONS

A fitted power law learning curve provides a forward-looking metric for analyzing prediction performance and can serve as a co-design tool to guide experimental biologists and computational scientists in the design of future experiments in prospective research studies.

Performance Comparison of Deep Learning Autoencoders for Cancer Subtype Detection Using Multi-Omics Data

A heterogeneous disease such as cancer is activated through multiple pathways and different perturbations. Depending upon the activated pathway(s), the survival of the patients varies significantly and shows different efficacy to various drugs. Therefore, cancer subtype detection using genomics level data is a significant research problem. Subtype detection is often a complex problem, and in most cases, needs multi-omics data fusion to achieve accurate subtyping. Different data fusion and subtyping approaches have been proposed over the years, such as kernel-based fusion, matrix factorization, and deep learning autoencoders. In this paper, we compared the performance of different deep learning autoencoders for cancer subtype detection. We performed cancer subtype detection on four different cancer types from The Cancer Genome Atlas (TCGA) datasets using four autoencoder implementations. We also predicted the optimal number of subtypes in a cancer type using the silhouette score and found that the detected subtypes exhibit significant differences in survival profiles. Furthermore, we compared the effect of feature selection and similarity measures for subtype detection. For further evaluation, we used the Glioblastoma multiforme (GBM) dataset and identified the differentially expressed genes in each of the subtypes. The results obtained are consistent with other genomic studies and can be corroborated with the involved pathways and biological functions. Thus, it shows that the results from the autoencoders, obtained through the interaction of different datatypes of cancer, can be used for the prediction and characterization of patient subgroups and survival profiles.

Predicting tumor response to drugs based on gene-expression biomarkers of sensitivity learned from cancer cell lines

Background

Human cancer cell line profiling and drug sensitivity studies provide valuable information about the therapeutic potential of drugs and their possible mechanisms of action. The goal of those studies is to translate the findings from in vitro studies of cancer cell lines into in vivo therapeutic relevance and, eventually, patients’ care. Tremendous progress has been made.

Results

In this work, we built predictive models for 453 drugs using data on gene expression and drug sensitivity (IC₅₀) from cancer cell lines. We identified many known drug-gene interactions and uncovered several potentially novel drug-gene associations. Importantly, we further applied these predictive models to ~ 17,000 bulk RNA-seq samples from The Cancer Genome Atlas (TCGA) and the Genotype-Tissue Expression (GTEx) database to predict drug sensitivity for both normal and tumor tissues. We created a web site for users to visualize and download our predicted data (https://manticore.niehs.nih.gov/cancerRxTissue). Using trametinib as an example, we showed that our approach can faithfully recapitulate the known tumor specificity of the drug.

Conclusions

We demonstrated that our approach can predict drugs that 1) are tumor-type specific; 2) elicit higher sensitivity from tumor compared to corresponding normal tissue; 3) elicit differential sensitivity across breast cancer subtypes. If validated, our prediction could have relevance for preclinical drug testing and in phase I clinical design.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-021-07581-7.

DeepDRK: a deep learning framework for drug repurposing through kernel-based multi-omics integration

Recent pharmacogenomic studies that generate sequencing data coupled with pharmacological characteristics for patient-derived cancer cell lines led to large amounts of multi-omics data for precision cancer medicine. Among various obstacles hindering clinical translation, lacking effective methods for multimodal and multisource data integration is becoming a bottleneck. Here we proposed DeepDRK, a machine learning framework for deciphering drug response through kernel-based data integration. To transfer information among different drugs and cancer types, we trained deep neural networks on more than 20 000 pan-cancer cell line-anticancer drug pairs. These pairs were characterized by kernel-based similarity matrices integrating multisource and multi-omics data including genomics, transcriptomics, epigenomics, chemical properties of compounds and known drug-target interactions. Applied to benchmark cancer cell line datasets, our model surpassed previous approaches with higher accuracy and better robustness. Then we applied our model on newly established patient-derived cancer cell lines and achieved satisfactory performance with AUC of 0.84 and AUPRC of 0.77. Moreover, DeepDRK was used to predict clinical response of cancer patients. Notably, the prediction of DeepDRK correlated well with clinical outcome of patients and revealed multiple drug repurposing candidates. In sum, DeepDRK provided a computational method to predict drug response of cancer cells from integrating pharmacogenomic datasets, offering an alternative way to prioritize repurposing drugs in precision cancer treatment. The DeepDRK is freely available via https://github.com/wangyc82/DeepDRK.

Artificial Intelligence in Cancer Research and Precision Medicine

Artificial intelligence (AI) is rapidly reshaping cancer research and personalized clinical care. Availability of high-dimensionality datasets coupled with advances in high-performance computing, as well as innovative deep learning architectures, has led to an explosion of AI use in various aspects of oncology research. These applications range from detection and classification of cancer, to molecular characterization of tumors and their microenvironment, to drug discovery and repurposing, to predicting treatment outcomes for patients. As these advances start penetrating the clinic, we foresee a shifting paradigm in cancer care becoming strongly driven by AI. SIGNIFICANCE: AI has the potential to dramatically affect nearly all aspects of oncology-from enhancing diagnosis to personalizing treatment and discovering novel anticancer drugs. Here, we review the recent enormous progress in the application of AI to oncology, highlight limitations and pitfalls, and chart a path for adoption of AI in the cancer clinic.

Clinical epigenetics settings for cancer and cardiovascular diseases: real-life applications of network medicine at the bedside

Despite impressive efforts invested in epigenetic research in the last 50 years, clinical applications are still lacking. Only a few university hospital centers currently use epigenetic biomarkers at the bedside. Moreover, the overall concept of precision medicine is not widely recognized in routine medical practice and the reductionist approach remains predominant in treating patients affected by major diseases such as cancer and cardiovascular diseases. By its' very nature, epigenetics is integrative of genetic networks. The study of epigenetic biomarkers has led to the identification of numerous drugs with an increasingly significant role in clinical therapy especially of cancer patients. Here, we provide an overview of clinical epigenetics within the context of network analysis. We illustrate achievements to date and discuss how we can move from traditional medicine into the era of network medicine (NM), where pathway-informed molecular diagnostics will allow treatment selection following the paradigm of precision medicine.

Drug ranking using machine learning systematically predicts the efficacy of anti-cancer drugs

Artificial intelligence and machine learning (ML) promise to transform cancer therapies by accurately predicting the most appropriate therapies to treat individual patients. Here, we present an approach, named Drug Ranking Using ML (DRUML), which uses omics data to produce ordered lists of >400 drugs based on their anti-proliferative efficacy in cancer cells. To reduce noise and increase predictive robustness, instead of individual features, DRUML uses internally normalized distance metrics of drug response as features for ML model generation. DRUML is trained using in-house proteomics and phosphoproteomics data derived from 48 cell lines, and it is verified with data comprised of 53 cellular models from 12 independent laboratories. We show that DRUML predicts drug responses in independent verification datasets with low error (mean squared error < 0.1 and mean Spearman's rank 0.7). In addition, we demonstrate that DRUML predictions of cytarabine sensitivity in clinical leukemia samples are prognostic of patient survival (Log rank p < 0.005). Our results indicate that DRUML accurately ranks anti-cancer drugs by their efficacy across a wide range of pathologies.

Incorporating Machine Learning into Established Bioinformatics Frameworks

The exponential growth of biomedical data in recent years has urged the application of numerous machine learning techniques to address emerging problems in biology and clinical research. By enabling the automatic feature extraction, selection, and generation of predictive models, these methods can be used to efficiently study complex biological systems. Machine learning techniques are frequently integrated with bioinformatic methods, as well as curated databases and biological networks, to enhance training and validation, identify the best interpretable features, and enable feature and model investigation. Here, we review recently developed methods that incorporate machine learning within the same framework with techniques from molecular evolution, protein structure analysis, systems biology, and disease genomics. We outline the challenges posed for machine learning, and, in particular, deep learning in biomedicine, and suggest unique opportunities for machine learning techniques integrated with established bioinformatics approaches to overcome some of these challenges.

Impact of between-tissue differences on pan-cancer predictions of drug sensitivity

Increased availability of drug response and genomics data for many tumor cell lines has accelerated the development of pan-cancer prediction models of drug response. However, it is unclear how much between-tissue differences in drug response and molecular characteristics may contribute to pan-cancer predictions. Also unknown is whether the performance of pan-cancer models could vary by cancer type. Here, we built a series of pan-cancer models using two datasets containing 346 and 504 cell lines, each with MEK inhibitor (MEKi) response and mRNA expression, point mutation, and copy number variation data, and found that, while the tissue-level drug responses are accurately predicted (between-tissue ρ = 0.88–0.98), only 5 of 10 cancer types showed successful within-tissue prediction performance (within-tissue ρ = 0.11–0.64). Between-tissue differences make substantial contributions to the performance of pan-cancer MEKi response predictions, as exclusion of between-tissue signals leads to a decrease in Spearman’s ρ from a range of 0.43–0.62 to 0.30–0.51. In practice, joint analysis of multiple cancer types usually has a larger sample size, hence greater power, than for one cancer type; and we observe that higher accuracy of pan-cancer prediction of MEKi response is almost entirely due to the sample size advantage. Success of pan-cancer prediction reveals how drug response in different cancers may invoke shared regulatory mechanisms despite tissue-specific routes of oncogenesis, yet predictions in different cancer types require flexible incorporation of between-cancer and within-cancer signals. As most datasets in genome sciences contain multiple levels of heterogeneity, careful parsing of group characteristics and within-group, individual variation is essential when making robust inference.

One of the central goals for precision oncology is to tailor treatment of individual tumors by their molecular characteristics. While drug response predictions have traditionally been sought within each cancer type, it has long been hoped to develop more robust predictions by jointly considering diverse cancer types. While such pan-cancer approaches have improved in recent years, it remains unclear whether between-tissue differences are contributing to the reported pan-cancer prediction performance. This concern stems from the observation that, when cancer types differ in both molecular features and drug response, strong predictive information can come mainly from differences among tissue types. Our study finds that both between- and within-cancer type signals provide substantial contributions to pan-cancer drug response prediction models, and about half of the cancer types examined are poorly predicted despite strong overall performance across all cancer types. We also find that pan-cancer prediction models perform similarly or better than cancer type-specific models, and in many cases the advantage of pan-cancer models is due to the larger number of samples available for pan-cancer analysis. Our results highlight tissue-of-origin as a key consideration for pan-cancer drug response prediction models, and recommend cancer type-specific considerations when translating pan-cancer prediction models for clinical use.

Anticancer drug synergy prediction in understudied tissues using transfer learning

OBJECTIVE

Drug combination screening has advantages in identifying cancer treatment options with higher efficacy without degradation in terms of safety. A key challenge is that the accumulated number of observations in in-vitro drug responses varies greatly among different cancer types, where some tissues are more understudied than the others. Thus, we aim to develop a drug synergy prediction model for understudied tissues as a way of overcoming data scarcity problems.

MATERIALS AND METHODS

We collected a comprehensive set of genetic, molecular, phenotypic features for cancer cell lines. We developed a drug synergy prediction model based on multitask deep neural networks to integrate multimodal input and multiple output. We also utilized transfer learning from data-rich tissues to data-poor tissues.

RESULTS

We showed improved accuracy in predicting synergy in both data-rich tissues and understudied tissues. In data-rich tissue, the prediction model accuracy was 0.9577 AUROC for binarized classification task and 174.3 mean squared error for regression task. We observed that an adequate transfer learning strategy significantly increases accuracy in the understudied tissues.

CONCLUSIONS

Our synergy prediction model can be used to rank synergistic drug combinations in understudied tissues and thus help to prioritize future in-vitro experiments. Code is available at https://github.com/yejinjkim/synergy-transfer.

Processing and Analysis of RNA-seq Data from Public Resources

Advances in next generation sequencing (NGS) technologies resulted in a broad array of large-scale gene expression studies and an unprecedented volume of whole messenger RNA (mRNA) sequencing data, or the transcriptome (also known as RNA sequencing, or RNA-seq). These include the Genotype Tissue Expression project (GTEx) and The Cancer Genome Atlas (TCGA), among others. Here we cover some of the commonly used datasets, provide an overview on how to begin the analysis pipeline, and how to explore and interpret the data provided by these publicly available resources.

Machine and deep learning approaches for cancer drug repurposing

Knowledge of the underpinnings of cancer initiation, progression and metastasis has increased exponentially in recent years. Advanced "omics" coupled with machine learning and artificial intelligence (deep learning) methods have helped elucidate targets and pathways critical to those processes that may be amenable to pharmacologic modulation. However, the current anti-cancer therapeutic armamentarium continues to lag behind. As the cost of developing a new drug remains prohibitively expensive, repurposing of existing approved and investigational drugs is sought after given known safety profiles and reduction in the cost barrier. Notably, successes in oncologic drug repurposing have been infrequent. Computational in-silico strategies have been developed to aid in modeling biological processes to find new disease-relevant targets and discovering novel drug-target and drug-phenotype associations. Machine and deep learning methods have especially enabled leaps in those successes. This review will discuss these methods as they pertain to cancer biology as well as immunomodulation for drug repurposing opportunities in oncologic diseases.

Network-based drug sensitivity prediction

Background

Drug sensitivity prediction and drug responsive biomarker selection on high-throughput genomic data is a critical step in drug discovery. Many computational methods have been developed to serve this purpose including several deep neural network models. However, the modular relations among genomic features have been largely ignored in these methods. To overcome this limitation, the role of the gene co-expression network on drug sensitivity prediction is investigated in this study.

Methods

In this paper, we first introduce a network-based method to identify representative features for drug response prediction by using the gene co-expression network. Then, two graph-based neural network models are proposed and both models integrate gene network information directly into neural network for outcome prediction. Next, we present a large-scale comparative study among the proposed network-based methods, canonical prediction algorithms (i.e., Elastic Net, Random Forest, Partial Least Squares Regression, and Support Vector Regression), and deep neural network models for drug sensitivity prediction. All the source code and processed datasets in this study are available at https://github.com/compbiolabucf/drug-sensitivity-prediction.

Results

In the comparison of different feature selection methods and prediction methods on a non-small cell lung cancer (NSCLC) cell line RNA-seq gene expression dataset with 50 different drug treatments, we found that (1) the network-based feature selection method improves the prediction performance compared to Pearson correlation coefficients; (2) Random Forest outperforms all the other canonical prediction algorithms and deep neural network models; (3) the proposed graph-based neural network models show better prediction performance compared to deep neural network model; (4) the prediction performance is drug dependent and it may relate to the drug’s mechanism of action.

Conclusions

Network-based feature selection method and prediction models improve the performance of the drug response prediction. The relations between the genomic features are more robust and stable compared to the correlation between each individual genomic feature and the drug response in high dimension and low sample size genomic datasets.

Deep learning of pharmacogenomics resources: moving towards precision oncology

The recent accumulation of cancer genomic data provides an opportunity to understand how a tumor's genomic characteristics can affect its responses to drugs. This field, called pharmacogenomics, is a key area in the development of precision oncology. Deep learning (DL) methodology has emerged as a powerful technique to characterize and learn from rapidly accumulating pharmacogenomics data. We introduce the fundamentals and typical model architectures of DL. We review the use of DL in classification of cancers and cancer subtypes (diagnosis and treatment stratification of patients), prediction of drug response and drug synergy for individual tumors (treatment prioritization for a patient), drug repositioning and discovery and the study of mechanism/mode of action of treatments. For each topic, we summarize current genomics and pharmacogenomics data resources such as pan-cancer genomics data for cancer cell lines (CCLs) and tumors, and systematic pharmacologic screens of CCLs. By revisiting the published literature, including our in-house analyses, we demonstrate the unprecedented capability of DL enabled by rapid accumulation of data resources to decipher complex drug response patterns, thus potentially improving cancer medicine. Overall, this review provides an in-depth summary of state-of-the-art DL methods and up-to-date pharmacogenomics resources and future opportunities and challenges to realize the goal of precision oncology.

Methods and resources to access mutation-dependent effects on cancer drug treatment

In clinical cancer treatment, genomic alterations would often affect the response of patients to anticancer drugs. Studies have shown that molecular features of tumors could be biomarkers predictive of sensitivity or resistance to anticancer agents, but the identification of actionable mutations are often constrained by the incomplete understanding of cancer genomes. Recent progresses of next-generation sequencing technology greatly facilitate the extensive molecular characterization of tumors and promote precision medicine in cancers. More and more clinical studies, cancer cell lines studies, CRISPR screening studies as well as patient-derived model studies were performed to identify potential actionable mutations predictive of drug response, which provide rich resources of molecularly and pharmacologically profiled cancer samples at different levels. Such abundance of data also enables the development of various computational models and algorithms to solve the problem of drug sensitivity prediction, biomarker identification and in silico drug prioritization by the integration of multiomics data. Here, we review the recent development of methods and resources that identifies mutation-dependent effects for cancer treatment in clinical studies, functional genomics studies and computational studies and discuss the remaining gaps and future directions in this area.

Single-Cell Techniques and Deep Learning in Predicting Drug Response

Rapidly developing single-cell sequencing analyses produce more comprehensive profiles of the genomic, transcriptomic, and epigenomic heterogeneity of tumor subpopulations than do traditional bulk sequencing analyses. Moreover, single-cell techniques allow the response of a tumor to drug exposure to be more thoroughlyinvestigated. Deep learning (DL) models have successfully extracted features from complex bulk sequence data to predict drug responses. We review recent innovations in single-cell technologies and DL-based approaches related to drug sensitivity predictions. We believe that, by using insights from bulk sequencedata, deep transfer learning (DTL) can facilitate the use of single-cell data for training superior DL-based drug prediction models.

Quantitative Structure-Mutation-Activity Relationship Tests (QSMART) model for protein kinase inhibitor response prediction

BACKGROUND

Protein kinases are a large family of druggable proteins that are genomically and proteomically altered in many human cancers. Kinase-targeted drugs are emerging as promising avenues for personalized medicine because of the differential response shown by altered kinases to drug treatment in patients and cell-based assays. However, an incomplete understanding of the relationships connecting genome, proteome and drug sensitivity profiles present a major bottleneck in targeting kinases for personalized medicine.

RESULTS

In this study, we propose a multi-component Quantitative Structure-Mutation-Activity Relationship Tests (QSMART) model and neural networks framework for providing explainable models of protein kinase inhibition and drug response ([Formula: see text]) profiles in cell lines. Using non-small cell lung cancer as a case study, we show that interaction terms that capture associations between drugs, pathways, and mutant kinases quantitatively contribute to the response of two EGFR inhibitors (afatinib and lapatinib). In particular, protein-protein interactions associated with the JNK apoptotic pathway, associations between lung development and axon extension, and interaction terms connecting drug substructures and the volume/charge of mutant residues at specific structural locations contribute significantly to the observed [Formula: see text] values in cell-based assays.

CONCLUSIONS

By integrating multi-omics data in the QSMART model, we not only predict drug responses in cancer cell lines with high accuracy but also identify features and explainable interaction terms contributing to the accuracy. Although we have tested our multi-component explainable framework on protein kinase inhibitors, it can be extended across the proteome to investigate the complex relationships connecting genotypes and drug sensitivity profiles.

Predicting Drug Response and Synergy Using a Deep Learning Model of Human Cancer Cells

Most drugs entering clinical trials fail, often related to an incomplete understanding of the mechanisms governing drug response. Machine learning techniques hold immense promise for better drug response predictions, but most have not reached clinical practice due to their lack of interpretability and their focus on monotherapies. We address these challenges by developing DrugCell, an interpretable deep learning model of human cancer cells trained on the responses of 1,235 tumor cell lines to 684 drugs. Tumor genotypes induce states in cellular subsystems that are integrated with drug structure to predict response to therapy and, simultaneously, learn biological mechanisms underlying the drug response. DrugCell predictions are accurate in cell lines and also stratify clinical outcomes. Analysis of DrugCell mechanisms leads directly to the design of synergistic drug combinations, which we validate systematically by combinatorial CRISPR, drug-drug screening in vitro, and patient-derived xenografts. DrugCell provides a blueprint for constructing interpretable models for predictive medicine.

Patient-derived cell line, xenograft and organoid models in lung cancer therapy

Lung cancer accounts for most cancer-related deaths worldwide and has an overall 5-year survival rate of ~15%. Cell lines have played important roles in the study of cancer biology and potential therapeutic targets, as well as pre-clinical testing of novel drugs. However, most experimental therapies that have cleared preclinical testing using established cell lines have failed phase III clinical trials. This suggests that such models may not adequately recapitulate patient tumor biology and clinical outcome predictions. Here, we discuss and compare different pre-clinical lung cancer models, including established cell lines, patient-derived cell lines, xenografts and organoids, summarize the methodology for generating these models, and review their relative advantages and limitations in different oncologic research applications. We further discuss additional gaps in patient-derived pre-clinical models to better recapitulate tumor biology and improve their clinical predictive power.

Leveraging TCGA gene expression data to build predictive models for cancer drug response

BACKGROUND

Machine learning has been utilized to predict cancer drug response from multi-omics data generated from sensitivities of cancer cell lines to different therapeutic compounds. Here, we build machine learning models using gene expression data from patients' primary tumor tissues to predict whether a patient will respond positively or negatively to two chemotherapeutics: 5-Fluorouracil and Gemcitabine.

RESULTS

We focused on 5-Fluorouracil and Gemcitabine because based on our exclusion criteria, they provide the largest numbers of patients within TCGA. Normalized gene expression data were clustered and used as the input features for the study. We used matching clinical trial data to ascertain the response of these patients via multiple classification methods. Multiple clustering and classification methods were compared for prediction accuracy of drug response. Clara and random forest were found to be the best clustering and classification methods, respectively. The results show our models predict with up to 86% accuracy; despite the study's limitation of sample size. We also found the genes most informative for predicting drug response were enriched in well-known cancer signaling pathways and highlighted their potential significance in chemotherapy prognosis.

CONCLUSIONS

Primary tumor gene expression is a good predictor of cancer drug response. Investment in larger datasets containing both patient gene expression and drug response is needed to support future work of machine learning models. Ultimately, such predictive models may aid oncologists with making critical treatment decisions.

Clinical trial design: Past, present, and future in the context of big data and precision medicine

Clinical trials are fundamental for advances in cancer treatment. The traditional framework of phase 1 to 3 trials is designed for incremental advances between regimens. However, our ability to understand and treat cancer has evolved with the increase in drugs targeting an expanding array of therapeutic targets, the development of progressively comprehensive data sets, and emerging computational analytics, all of which are reshaping our treatment strategies. A more robust linkage between drugs and underlying cancer biology is blurring historical lines that define trials on the basis of cancer type. The complexity of the molecular basis of cancer, coupled with manifold variations in clinical status, is driving the individually tailored use of combinations of precision targeted drugs. This approach is spawning a new era of clinical trial types. Although most care is delivered in a community setting, large centers support real‐time multi‐omic analytics and their integrated interpretation by using machine learning in the context of real‐world data sets. Coupling the analytic capabilities of large centers to the tailored delivery of therapy in the community is forging a paradigm that is optimizing service for patients. Understanding the importance of these evolving trends across the health care spectrum will affect our treatment of cancer in the future and is the focus of this review.

With advances in cancer biology, precision therapeutics, and big data, clinical trial designs are evolving. They are transforming cancer care and research across the biomedical enterprise.

ADRML: anticancer drug response prediction using manifold learning

One of the prominent challenges in precision medicine is to select the most appropriate treatment strategy for each patient based on the personalized information. The availability of massive data about drugs and cell lines facilitates the possibility of proposing efficient computational models for predicting anticancer drug response. In this study, we propose ADRML, a model for Anticancer Drug Response Prediction using Manifold Learning to systematically integrate the cell line information with the drug information to make accurate predictions about drug therapeutic. The proposed model maps the drug response matrix into the lower-rank spaces that lead to obtaining new perspectives about cell lines and drugs. The drug response for a new cell line-drug pair is computed using the low-rank features. The evaluation of ADRML performance on various types of cell lines and drug information, in addition to the comparisons with previously proposed methods, shows that ADRML provides accurate and robust predictions. Further investigations about the association between drug response and pathway activity scores reveal that the predicted drug responses can shed light on the underlying drug mechanism. Also, the case studies suggest that the predictions of ADRML about novel cell line-drug pairs are validated by reliable pieces of evidence from the literature. Consequently, the evaluations verify that ADRML can be used in accurately predicting and imputing the anticancer drug response.

Computational Prediction of Drug Responses in Cancer Cell Lines From Cancer Omics and Detection of Drug Effectiveness Related Methylation Sites

Accurately predicting the response of a cancer patient to a therapeutic agent remains an important challenge in precision medicine. With the rise of data science, researchers have applied computational models to study the drug inhibition effects on cancers based on cancer genomics and transcriptomics. Moreover, a common epigenetic modification, DNA methylation, has been related to the occurrence and development of cancer, as well as drug effectiveness. Therefore, it is helpful for improvement of drug response prediction through exploring the relationship between DNA methylation and drug effectiveness. Here, we proposed a computational model to predict drug responses in cancers through integration of cancer genomics, transcriptomics, epigenomics, and compound chemical properties. Meanwhile, we applied a regularized regression model (Least Absolute Shrinkage and Selection Operator, lasso) to detect the methylation sites that were closely related to drug effectiveness. The prediction models were trained on a well-known pharmacogenomics data resource, Genomics of Drug Sensitivity in Cancer (GDSC). The cross-validation indicates that the performance of the prediction model using DNA methylation is comparable to that of using other cancer omics, including oncogene mutation and gene expression data. It indicates the important role of DNA methylation in prediction of drug responses. Encyclopedia of DNA Elements (ENCODE) and Transcriptional Regulatory Relationships Unraveled by Sentence-based Text mining (TRRUST2) database analyses suggest that the methylation sites associated with drug effectiveness are mainly located in the transcription factor (TF) binding region. Therefore, we hypothesized that the sensitivity of cancer cells to drugs could be regulated by changing the methylation modification of TF binding region. In conclusion, we confirmed the important role of DNA methylation in prediction of drug responses, and provided some methylation sites that closely related to the drug effectiveness, which may be a great regulatory target for improvement of drug treatment effects on cancer patients.

Classification of Cancer Types Using Graph Convolutional Neural Networks

BACKGROUND

Cancer has been a leading cause of death in the United States with significant health care costs. Accurate prediction of cancers at an early stage and understanding the genomic mechanisms that drive cancer development are vital to the improvement of treatment outcomes and survival rates, thus resulting in significant social and economic impacts. Attempts have been made to classify cancer types with machine learning techniques during the past two decades and deep learning approaches more recently.

RESULTS

In this paper, we established four models with graph convolutional neural network (GCNN) that use unstructured gene expressions as inputs to classify different tumor and non-tumor samples into their designated 33 cancer types or as normal. Four GCNN models based on a co-expression graph, co-expression+singleton graph, protein-protein interaction (PPI) graph, and PPI+singleton graph have been designed and implemented. They were trained and tested on combined 10,340 cancer samples and 731 normal tissue samples from The Cancer Genome Atlas (TCGA) dataset. The established GCNN models achieved excellent prediction accuracies (89.9-94.7%) among 34 classes (33 cancer types and a normal group). In silico gene-perturbation experiments were performed on four models based on co-expression graph, co-expression+singleton, PPI graph, and PPI+singleton graphs. The co-expression GCNN model was further interpreted to identify a total of 428 markers genes that drive the classification of 33 cancer types and normal. The concordance of differential expressions of these markers between the represented cancer type and others are confirmed. Successful classification of cancer types and a normal group regardless of normal tissues' origin suggested that the identified markers are cancer-specific rather than tissue-specific.

CONCLUSION

Novel GCNN models have been established to predict cancer types or normal tissue based on gene expression profiles. We demonstrated the results from the TCGA dataset that these models can produce accurate classification (above 94%), using cancer-specific markers genes. The models and the source codes are publicly available and can be readily adapted to the diagnosis of cancer and other diseases by the data-driven modeling research community.

Machine Learning in oncology: A clinical appraisal

Machine learning (ML) is a branch of artificial intelligence centered on algorithms which do not need explicit prior programming to function but automatically learn from available data, creating decision models to complete tasks. ML-based tools have numerous promising applications in several fields of medicine. Its use has grown following the increased availability of patient data due to technological advances such as digital health records and high-volume information extraction from medical images. Multiple ML algorithms have been proposed for applications in oncology. For instance, they have been employed for oncological risk assessment, automated segmentation, lesion detection, characterization, grading and staging, prediction of prognosis and therapy response. In the near future, ML could become essential part of every step of oncological screening strategies and patients' management thus leading to precision medicine.

Convolutional neural network models for cancer type prediction based on gene expression

BACKGROUND

Precise prediction of cancer types is vital for cancer diagnosis and therapy. Through a predictive model, important cancer marker genes can be inferred. Several studies have attempted to build machine learning models for this task however none has taken into consideration the effects of tissue of origin that can potentially bias the identification of cancer markers.

RESULTS

In this paper, we introduced several Convolutional Neural Network (CNN) models that take unstructured gene expression inputs to classify tumor and non-tumor samples into their designated cancer types or as normal. Based on different designs of gene embeddings and convolution schemes, we implemented three CNN models: 1D-CNN, 2D-Vanilla-CNN, and 2D-Hybrid-CNN. The models were trained and tested on gene expression profiles from combined 10,340 samples of 33 cancer types and 713 matched normal tissues of The Cancer Genome Atlas (TCGA). Our models achieved excellent prediction accuracies (93.9-95.0%) among 34 classes (33 cancers and normal). Furthermore, we interpreted one of the models, 1D-CNN model, with a guided saliency technique and identified a total of 2090 cancer markers (108 per class on average). The concordance of differential expression of these markers between the cancer type they represent and others is confirmed. In breast cancer, for instance, our model identified well-known markers, such as GATA3 and ESR1. Finally, we extended the 1D-CNN model for the prediction of breast cancer subtypes and achieved an average accuracy of 88.42% among 5 subtypes. The codes can be found at https://github.com/chenlabgccri/CancerTypePrediction.

CONCLUSIONS

Here we present novel CNN designs for accurate and simultaneous cancer/normal and cancer types prediction based on gene expression profiles, and unique model interpretation scheme to elucidate biologically relevance of cancer marker genes after eliminating the effects of tissue-of-origin. The proposed model has light hyperparameters to be trained and thus can be easily adapted to facilitate cancer diagnosis in the future.

How Machine Learning Will Transform Biomedicine

This Perspective explores the application of machine learning toward improved diagnosis and treatment. We outline a vision for how machine learning can transform three broad areas of biomedicine: clinical diagnostics, precision treatments, and health monitoring, where the goal is to maintain health through a range of diseases and the normal aging process. For each area, early instances of successful machine learning applications are discussed, as well as opportunities and challenges for machine learning. When these challenges are met, machine learning promises a future of rigorous, outcomes-based medicine with detection, diagnosis, and treatment strategies that are continuously adapted to individual and environmental differences.

Computational Methods for the Integrative Analysis of Genomics and Pharmacological Data

Since the pioneering NCI-60 panel of the late'80's, several major screenings of genetic profiling and drug testing in cancer cell lines have been conducted to investigate how genetic backgrounds and transcriptional patterns shape cancer's response to therapy and to identify disease-specific genes associated with drug response. Historically, pharmacogenomics screenings have been largely heterogeneous in terms of investigated cell lines, assay technologies, number of compounds, type and quality of genomic data, and methods for their computational analysis. The analysis of this enormous and heterogeneous amount of data required the development of computational methods for the integration of genomic profiles with drug responses across multiple screenings. Here, we will review the computational tools that have been developed to integrate cancer cell lines' genomic profiles and sensitivity to small molecule perturbations obtained from different screenings.

Zebrafish Avatars towards Personalized Medicine-A Comparative Review between Avatar Models

Cancer frequency and prevalence have been increasing in the past decades, with devastating impacts on patients and their families. Despite the great advances in targeted approaches, there is still a lack of methods to predict individual patient responses, and therefore treatments are tailored according to average response rates. “Omics” approaches are used for patient stratification and choice of therapeutic options towards a more precise medicine. These methods, however, do not consider all genetic and non-genetic dynamic interactions that occur upon drug treatment. Therefore, the need to directly challenge patient cells in a personalized manner remains. The present review addresses the state of the art of patient-derived in vitro and in vivo models, from organoids to mouse and zebrafish Avatars. The predictive power of each model based on the retrospective correlation with the patient clinical outcome will be considered. Finally, the review is focused on the emerging zebrafish Avatars and their unique characteristics allowing a fast analysis of local and systemic effects of drug treatments at the single-cell level. We also address the technical challenges that the field has yet to overcome.