The life-saving benefit of dexamethasone in severe COVID-19 is linked to a reversal of monocyte dysregulation.

Dexamethasone is a life-saving treatment for severe COVID-19, yet its mechanism of action is unknown, and many patients deteriorate or die despite timely treatment initiation. Here, we identify dexamethasone treatment-induced cellular and molecular changes associated with improved survival in COVID-19 patients. We observed a reversal of transcriptional hallmark signatures in monocytes associated with severe COVID-19 and the induction of a monocyte substate characterized by the expression of glucocorticoid-response genes. These molecular responses to dexamethasone were detected in circulating and pulmonary monocytes, and they were directly linked to survival. Monocyte single-cell RNA sequencing (scRNA-seq)-derived signatures were enriched in whole blood transcriptomes of patients with fatal outcome in two independent cohorts, highlighting the potential for identifying non-responders refractory to dexamethasone. Our findings link the effects of dexamethasone to specific immunomodulation and reversal of monocyte dysregulation, and they highlight the potential of single-cell omics for monitoring in vivo target engagement of immunomodulatory drugs and for patient stratification for precision medicine approaches.

Advancing drug-response prediction using multi-modal and -omics machine learning integration (MOMLIN): a case study on breast cancer clinical data.

The inherent heterogeneity of cancer contributes to highly variable responses to any anticancer treatments. This underscores the need to first identify precise biomarkers through complex multi-omics datasets that are now available. Although much research has focused on this aspect, identifying biomarkers associated with distinct drug responders still remains a major challenge. Here, we develop MOMLIN, a multi-modal and -omics machine learning integration framework, to enhance drug-response prediction. MOMLIN jointly utilizes sparse correlation algorithms and class-specific feature selection algorithms, which identifies multi-modal and -omics-associated interpretable components. MOMLIN was applied to 147 patients' breast cancer datasets (clinical, mutation, gene expression, tumor microenvironment cells and molecular pathways) to analyze drug-response class predictions for non-responders and variable responders. Notably, MOMLIN achieves an average AUC of 0.989, which is at least 10% greater when compared with current state-of-the-art (data integration analysis for biomarker discovery using latent components, multi-omics factor analysis, sparse canonical correlation analysis). Moreover, MOMLIN not only detects known individual biomarkers such as genes at mutation/expression level, most importantly, it correlates multi-modal and -omics network biomarkers for each response class. For example, an interaction between ER-negative-HMCN1-COL5A1 mutations-FBXO2-CSF3R expression-CD8 emerge as a multimodal biomarker for responders, potentially affecting antimicrobial peptides and FLT3 signaling pathways. In contrast, for resistance cases, a distinct combination of lymph node-TP53 mutation-PON3-ENSG00000261116 lncRNA expression-HLA-E-T-cell exclusions emerged as multimodal biomarkers, possibly impacting neurotransmitter release cycle pathway. MOMLIN, therefore, is expected advance precision medicine, such as to detect context-specific multi-omics network biomarkers and better predict drug-response classifications.

Optimal fusion of genotype and drug embeddings in predicting cancer drug response.

Predicting cancer drug response using both genomics and drug features has shown some success compared to using genomics features alone. However, there has been limited research done on how best to combine or fuse the two types of features. Using a visible neural network with two deep learning branches for genes and drug features as the base architecture, we experimented with different fusion functions and fusion points. Our experiments show that injecting multiplicative relationships between gene and drug latent features into the original concatenation-based architecture DrugCell significantly improved the overall predictive performance and outperformed other baseline models. We also show that different fusion methods respond differently to different fusion points, indicating that the relationship between drug features and different hierarchical biological level of gene features is optimally captured using different methods. Considering both predictive performance and runtime speed, tensor product partial is the best-performing fusion function to combine late-stage representations of drug and gene features to predict cancer drug response.

Computational drug discovery for castration-resistant prostate cancers through in vitro drug response modeling.

Prostate cancer (PC) is the most frequently diagnosed malignancy and a leading cause of cancer deaths in US men. Many PC cases metastasize and develop resistance to systemic hormonal therapy, a stage known as castration-resistant prostate cancer (CRPC). Therefore, there is an urgent need to develop effective therapeutic strategies for CRPC. Traditional drug discovery pipelines require significant time and capital input, which highlights a need for novel methods to evaluate the repositioning potential of existing drugs. Here, we present a computational framework to predict drug sensitivities of clinical CRPC tumors to various existing compounds and identify treatment options with high potential for clinical impact. We applied this method to a CRPC patient cohort and nominated drugs to combat resistance to hormonal therapies including abiraterone and enzalutamide. The utility of this method was demonstrated by nomination of multiple drugs that are currently undergoing clinical trials for CRPC. Additionally, this method identified the tetracycline derivative COL-3, for which we validated higher efficacy in an isogenic cell line model of enzalutamide-resistant vs. enzalutamide-sensitive CRPC. In enzalutamide-resistant CRPC cells, COL-3 displayed higher activity for inhibiting cell growth and migration, and for inducing G1-phase cell cycle arrest and apoptosis. Collectively, these findings demonstrate the utility of a computational framework for independent validation of drugs being tested in CRPC clinical trials, and for nominating drugs with enhanced biological activity in models of enzalutamide-resistant CRPC. The efficiency of this method relative to traditional drug development approaches indicates a high potential for accelerating drug development for CRPC.

Functional and clinical roles of stromal PDGF receptors in tumor biology.

PDGF receptors play pivotal roles in both developmental and physiological processes through the regulation of mesenchymal cells involved in paracrine instructive interactions with epithelial or endothelial cells. Tumor biology studies, alongside analyses of patient tissue samples, provide strong indications that the PDGF signaling pathways are also critical in various types of human cancer. This review summarizes experimental findings and correlative studies, which have explored the biological mechanisms and clinical relevance of PDGFRs in mesenchymal cells of the tumor microenvironment. Collectively, these studies support the overall concept that the PDGF system is a critical regulator of tumor growth, metastasis, and drug efficacy, suggesting yet unexploited targeting opportunities. The inter-patient variability in stromal PDGFR expression, as being linked to prognosis and treatment responses, not only indicates the need for stratified approaches in upcoming therapeutic investigations but also implies the potential for the development of PDGFRs as biomarkers of clinical utility, interestingly also in settings outside PDGFR-directed treatments.

iBT-Net: an incremental broad transformer network for cancer drug response prediction.

In modern precision medicine, it is an important research topic to predict cancer drug response. Due to incomplete chemical structures and complex gene features, however, it is an ongoing work to design efficient data-driven methods for predicting drug response. Moreover, since the clinical data cannot be easily obtained all at once, the data-driven methods may require relearning when new data are available, resulting in increased time consumption and cost. To address these issues, an incremental broad Transformer network (iBT-Net) is proposed for cancer drug response prediction. Different from the gene expression features learning from cancer cell lines, structural features are further extracted from drugs by Transformer. Broad learning system is then designed to integrate the learned gene features and structural features of drugs to predict the response. With the capability of incremental learning, the proposed method can further use new data to improve its prediction performance without retraining totally. Experiments and comparison studies demonstrate the effectiveness and superiority of iBT-Net under different experimental configurations and continuous data learning.

GADRP: graph convolutional networks and autoencoders for cancer drug response prediction.

Drug response prediction in cancer cell lines is of great significance in personalized medicine. In this study, we propose GADRP, a cancer drug response prediction model based on graph convolutional networks (GCNs) and autoencoders (AEs). We first use a stacked deep AE to extract low-dimensional representations from cell line features, and then construct a sparse drug cell line pair (DCP) network incorporating drug, cell line, and DCP similarity information. Later, initial residual and layer attention-based GCN (ILGCN) that can alleviate over-smoothing problem is utilized to learn DCP features. And finally, fully connected network is employed to make prediction. Benchmarking results demonstrate that GADRP can significantly improve prediction performance on all metrics compared with baselines on five datasets. Particularly, experiments of predictions of unknown DCP responses, drug-cancer tissue associations， and drug-pathway associations illustrate the predictive power of GADRP. All results highlight the effectiveness of GADRP in predicting drug responses, and its potential value in guiding anti-cancer drug selection.

A systematic assessment of deep learning methods for drug response prediction: from in vitro to clinical applications.

Drug response prediction is an important problem in personalized cancer therapy. Among various newly developed models, significant improvement in prediction performance has been reported using deep learning methods. However, systematic comparisons of deep learning methods, especially of the transferability from preclinical models to clinical cohorts, are currently lacking. To provide a more rigorous assessment, the performance of six representative deep learning methods for drug response prediction using nine evaluation metrics, including the overall prediction accuracy, predictability of each drug, potential associated factors and transferability to clinical cohorts, in multiple application scenarios was benchmarked. Most methods show promising prediction within cell line datasets, and TGSA, with its lower time cost and better performance, is recommended. Although the performance metrics decrease when applying models trained on cell lines to patients, a certain amount of power to distinguish clinical response on some drugs can be maintained using CRDNN and TGSA. With these assessments, we provide a guidance for researchers to choose appropriate methods, as well as insights into future directions for the development of more effective methods in clinical scenarios.

Immunoinformatic Approach for Rational Identification of Immunogenic Peptides Against Host Entry and/or Exit Mpox Proteins and Potential Multiepitope Vaccine Construction.

COVID-19 has intensified humanity's concern about the emergence of new pandemics. Since 2018, epidemic outbreaks of the mpox virus have become worrisome. In June 2022, the World Health Organization declared the disease a global health emergency, with 14 500 cases reported by the Centers for Disease Control and Prevention in 60 countries. Therefore, the development of a vaccine based on the current virus genome is paramount in combating new cases. In view of this, we hypothesized the obtainment of rational immunogenic peptides predicted from proteins responsible for entry of the mpox virus into the host (A17L, A26L/A30L, A33R, H2R, L1R), exit (A27L, A35R, A36R, C19L), and both (B5R). To achieve this, we aligned the genome sequencing data of mpox virus isolated from an infected individual in the United States in June 2022 (ON674051.1) with the reference genome dated 2001 (NC_003310.1) for conservation analysis. The Immune Epitope Database server was used for the identification and characterization of the epitopes of each protein related to major histocompatibility complex I or II interaction and recognition by B-cell receptors, resulting in 138 epitopes for A17L, 233 for A28L, 48 for A33R, 77 for H2R, 77 for L1R, 270 for A27L, 72 for A35R, A36R, 148 for C19L, and 276 for B5R. These epitopes were tested in silico for antigenicity, physicochemical properties, and allergenicity, resulting in 51, 40, 10, 34, 38, 57, 25, 7, 47, and 53 epitopes, respectively. Additionally, to select an epitope with the highest promiscuity of binding to major histocompatibility complexes and B-cell receptor simultaneously, all epitopes of each protein were aligned, and the most repetitive and antigenic regions were identified. By classifying the results, we obtained 23 epitopes from the entry proteins, 16 from the exit proteins, and 7 from both. Subsequently, 1 epitope from each protein was selected, and all 3 were fused to construct a chimeric protein that has potential as a multiepitope vaccine. The constructed vaccine was then analyzed for its physicochemical, antigenic, and allergenic properties. Protein modeling, molecular dynamics, and molecular docking were performed on Toll-like receptors 2, 4, and 8, followed by in silico immune simulation of the vaccine. Finally, the results indicate an effective, stable, and safe vaccine that can be further tested, especially in vitro and in vivo, to validate the findings demonstrated in silico.

DeepAEG: a model for predicting cancer drug response based on data enhancement and edge-collaborative update strategies.

MOTIVATION: The prediction of cancer drug response is a challenging subject in modern personalized cancer therapy due to the uncertainty of drug efficacy and the heterogeneity of patients. It has been shown that the characteristics of the drug itself and the genomic characteristics of the patient can greatly influence the results of cancer drug response. Therefore, accurate, efficient, and comprehensive methods for drug feature extraction and genomics integration are crucial to improve the prediction accuracy. RESULTS: Accurate prediction of cancer drug response is vital for guiding the design of anticancer drugs. In this study, we propose an end-to-end deep learning model named DeepAEG which is based on a complete-graph update mode to predict IC50. Specifically, we integrate an edge update mechanism on the basis of a hybrid graph convolutional network to comprehensively learn the potential high-dimensional representation of topological structures in drugs, including atomic characteristics and chemical bond information. Additionally, we present a novel approach for enhancing simplified molecular input line entry specification data by employing sequence recombination to eliminate the defect of single sequence representation of drug molecules. Our extensive experiments show that DeepAEG outperforms other existing methods across multiple evaluation parameters in multiple test sets. Furthermore, we identify several potential anticancer agents, including bortezomib, which has proven to be an effective clinical treatment option. Our results highlight the potential value of DeepAEG in guiding the design of specific cancer treatment regimens.

iGenSig-Rx: an integral genomic signature based white-box tool for modeling cancer therapeutic responses using multi-omics data.

Multi-omics sequencing is poised to revolutionize clinical care in the coming decade. However, there is a lack of effective and interpretable genome-wide modeling methods for the rational selection of patients for personalized interventions. To address this, we present iGenSig-Rx, an integral genomic signature-based approach, as a transparent tool for modeling therapeutic response using clinical trial datasets. This method adeptly addresses challenges related to cross-dataset modeling by capitalizing on high-dimensional redundant genomic features, analogous to reinforcing building pillars with redundant steel rods. Moreover, it integrates adaptive penalization of feature redundancy on a per-sample basis to prevent score flattening and mitigate overfitting. We then developed a purpose-built R package to implement this method for modeling clinical trial datasets. When applied to genomic datasets for HER2 targeted therapies, iGenSig-Rx model demonstrates consistent and reliable predictive power across four independent clinical trials. More importantly, the iGenSig-Rx model offers the level of transparency much needed for clinical application, allowing for clear explanations as to how the predictions are produced, how the features contribute to the prediction, and what are the key underlying pathways. We anticipate that iGenSig-Rx, as an interpretable class of multi-omics modeling methods, will find broad applications in big-data based precision oncology. The R package is available: https://github.com/wangxlab/iGenSig-Rx .

Enhancing pathological complete response prediction in breast cancer: the role of dynamic characterization of DCE-MRI and its association with tumor heterogeneity.

BACKGROUND: Early prediction of pathological complete response (pCR) is important for deciding appropriate treatment strategies for patients. In this study, we aimed to quantify the dynamic characteristics of dynamic contrast-enhanced magnetic resonance images (DCE-MRI) and investigate its value to improve pCR prediction as well as its association with tumor heterogeneity in breast cancer patients. METHODS: The DCE-MRI, clinicopathologic record, and full transcriptomic data of 785 breast cancer patients receiving neoadjuvant chemotherapy were retrospectively included from a public dataset. Dynamic features of DCE-MRI were computed from extracted phase-varying radiomic feature series using 22 CAnonical Time-sereis CHaracteristics. Dynamic model and radiomic model were developed by logistic regression using dynamic features and traditional radiomic features respectively. Various combined models with clinical factors were also developed to find the optimal combination and the significance of each components was evaluated. All the models were evaluated in independent test set in terms of area under receiver operating characteristic curve (AUC). To explore the potential underlying biological mechanisms, radiogenomic analysis was implemented on patient subgroups stratified by dynamic model to identify differentially expressed genes (DEGs) and enriched pathways. RESULTS: A 10-feature dynamic model and a 4-feature radiomic model were developed (AUC = 0.688, 95%CI: 0.635-0.741 and AUC = 0.650, 95%CI: 0.595-0.705) and tested (AUC = 0.686, 95%CI: 0.594-0.778 and AUC = 0.626, 95%CI: 0.529-0.722), with the dynamic model showing slightly higher AUC (train p = 0.181, test p = 0.222). The combined model of clinical, radiomic, and dynamic achieved the highest AUC in pCR prediction (train: 0.769, 95%CI: 0.722-0.816 and test: 0.762, 95%CI: 0.679-0.845). Compared with clinical-radiomic combined model (train AUC = 0.716, 95%CI: 0.665-0.767 and test AUC = 0.695, 95%CI: 0.656-0.714), adding the dynamic component brought significant improvement in model performance (train p < 0.001 and test p = 0.005). Radiogenomic analysis identified 297 DEGs, including CXCL9, CCL18, and HLA-DPB1 which are known to be associated with breast cancer prognosis or angiogenesis. Gene set enrichment analysis further revealed enrichment of gene ontology terms and pathways related to immune system. CONCLUSION: Dynamic characteristics of DCE-MRI were quantified and used to develop dynamic model for improving pCR prediction in breast cancer patients. The dynamic model was associated with tumor heterogeniety in prognostic-related gene expression and immune-related pathways.

Stratifying esophago-gastric cancer treatment using a patient-derived organoid-based threshold.

BACKGROUND AND AIMS: This study sought to determine the value of patient-derived organoids (PDOs) from esophago-gastric adenocarcinoma (EGC) for response prediction to neoadjuvant chemotherapy (neoCTx). METHODS: Endoscopic biopsies of patients with locally advanced EGC (n = 120) were taken into culture and PDOs expanded. PDOs' response towards the single substances of the FLOT regimen and the combination treatment were correlated to patients' pathological response using tumor regression grading. A classifier based on FLOT response of PDOs was established in an exploratory cohort (n = 13) and subsequently confirmed in an independent validation cohort (n = 13). RESULTS: EGC PDOs reflected patients' diverse responses to single chemotherapeutics and the combination regimen FLOT. In the exploratory cohort, PDOs response to single 5-FU and FLOT combination treatment correlated with the patients' pathological response (5-FU: Kendall's τ = 0.411, P = 0.001; FLOT: Kendall's τ = 0.694, P = 2.541e-08). For FLOT testing, a high diagnostic precision in receiver operating characteristic (ROC) analysis was reached with an AUCROC of 0.994 (CI 0.980 to 1.000). The discriminative ability of PDO-based FLOT testing allowed the definition of a threshold, which classified in an independent validation cohort FLOT responders from non-responders with high sensitivity (90%), specificity (100%) and accuracy (92%). CONCLUSION: In vitro drug testing of EGC PDOs has a high predictive accuracy in classifying patients' histological response to neoadjuvant FLOT treatment. Taking into account the high rate of successful PDO expansion from biopsies, the definition of a threshold that allows treatment stratification paves the way for an interventional trial exploring PDO-guided treatment of EGC patients.

A prediction model for response to immune checkpoint inhibition in advanced melanoma.

Predicting who will benefit from treatment with immune checkpoint inhibition (ICI) in patients with advanced melanoma is challenging. We developed a multivariable prediction model for response to ICI, using routinely available clinical data including primary melanoma characteristics. We used a population-based cohort of 3525 patients with advanced cutaneous melanoma treated with anti-PD-1-based therapy. Our prediction model for predicting response within 6 months after ICI initiation was internally validated with bootstrap resampling. Performance evaluation included calibration, discrimination and internal-external cross-validation. Included patients received anti-PD-1 monotherapy (n = 2366) or ipilimumab plus nivolumab (n = 1159) in any treatment line. The model included serum lactate dehydrogenase, World Health Organization performance score, type and line of ICI, disease stage and time to first distant recurrence-all at start of ICI-, and location and type of primary melanoma, the presence of satellites and/or in-transit metastases at primary diagnosis and sex. The over-optimism adjusted area under the receiver operating characteristic was 0.66 (95% CI: 0.64-0.66). The range of predicted response probabilities was 7%-81%. Based on these probabilities, patients were categorized into quartiles. Compared to the lowest response quartile, patients in the highest quartile had a significantly longer median progression-free survival (20.0 vs 2.8 months; P < .001) and median overall survival (62.0 vs 8.0 months; P < .001). Our prediction model, based on routinely available clinical variables and primary melanoma characteristics, predicts response to ICI in patients with advanced melanoma and discriminates well between treated patients with a very good and very poor prognosis.

Clinical and genetic factors associated with tumor response to neoadjuvant (chemo)radiotherapy, survival and recurrence risk in rectal cancer.

Rectal cancer poses challenges in preoperative treatment response, with up to 30% achieving a complete response (CR). Personalized treatment relies on accurate identification of responders at diagnosis. This study aimed to unravel CR determinants, overall survival (OS), and time to recurrence (TTR) using clinical and targeted sequencing data. Analyzing 402 patients undergoing preoperative treatment, tumor stage, size, and treatment emerged as robust response predictors. CR rates were higher in smaller, early-stage, and intensively treated tumors. Targeted sequencing analyzed 216 cases, while 120 patients provided hotspot mutation data. KRAS mutation dramatically reduced CR odds by over 50% (odds ratio [OR] = 0.3 in the targeted sequencing and OR = 0.4 hotspot cohorts, respectively). In contrast, SMAD4 and SYNE1 mutations were associated with higher CR rates (OR = 6.0 and 6.8, respectively). Favorable OS was linked to younger age, CR, and low baseline carcinoembryonic antigen levels. Notably, CR and an APC mutation increased TTR, while a BRAF mutation negatively affected TTR. Beyond tumor burden, SMAD4 and SYNE1 mutations significantly influenced CR. KRAS mutations independently correlated with radiotherapy resistance, and BRAF mutations heightened recurrence risk. Intriguingly, non-responding tumors with initially small sizes carried a higher risk of recurrence. The findings, even if limited in addition to the imperfect clinical factors, offer insights into rectal cancer treatment response, guiding personalized therapeutic strategies. By uncovering factors impacting CR, OS, and TTR, this study underscores the importance of tailored approaches for rectal cancer patients. These findings, based on extensive analysis and mutation data, pave the way for personalized interventions, optimizing outcomes in the challenges of rectal cancer preoperative treatment.

A cross-study analysis of drug response prediction in cancer cell lines.

To enable personalized cancer treatment, machine learning models have been developed to predict drug response as a function of tumor and drug features. However, most algorithm development efforts have relied on cross-validation within a single study to assess model accuracy. While an essential first step, cross-validation within a biological data set typically provides an overly optimistic estimate of the prediction performance on independent test sets. To provide a more rigorous assessment of model generalizability between different studies, we use machine learning to analyze five publicly available cell line-based data sets: National Cancer Institute 60, ancer Therapeutics Response Portal (CTRP), Genomics of Drug Sensitivity in Cancer, Cancer Cell Line Encyclopedia and Genentech Cell Line Screening Initiative (gCSI). Based on observed experimental variability across studies, we explore estimates of prediction upper bounds. We report performance results of a variety of machine learning models, with a multitasking deep neural network achieving the best cross-study generalizability. By multiple measures, models trained on CTRP yield the most accurate predictions on the remaining testing data, and gCSI is the most predictable among the cell line data sets included in this study. With these experiments and further simulations on partial data, two lessons emerge: (1) differences in viability assays can limit model generalizability across studies and (2) drug diversity, more than tumor diversity, is crucial for raising model generalizability in preclinical screening.

DeepTTA: a transformer-based model for predicting cancer drug response.

Identifying new lead molecules to treat cancer requires more than a decade of dedicated effort. Before selected drug candidates are used in the clinic, their anti-cancer activity is generally validated by in vitro cellular experiments. Therefore, accurate prediction of cancer drug response is a critical and challenging task for anti-cancer drugs design and precision medicine. With the development of pharmacogenomics, the combination of efficient drug feature extraction methods and omics data has made it possible to use computational models to assist in drug response prediction. In this study, we propose DeepTTA, a novel end-to-end deep learning model that utilizes transformer for drug representation learning and a multilayer neural network for transcriptomic data prediction of the anti-cancer drug responses. Specifically, DeepTTA uses transcriptomic gene expression data and chemical substructures of drugs for drug response prediction. Compared to existing methods, DeepTTA achieved higher performance in terms of root mean square error, Pearson correlation coefficient and Spearman's rank correlation coefficient on multiple test sets. Moreover, we discovered that anti-cancer drugs bortezomib and dactinomycin provide a potential therapeutic option with multiple clinical indications. With its excellent performance, DeepTTA is expected to be an effective method in cancer drug design.

GraphCDR: a graph neural network method with contrastive learning for cancer drug response prediction.

Predicting the response of a cancer cell line to a therapeutic drug is an important topic in modern oncology that can help personalized treatment for cancers. Although numerous machine learning methods have been developed for cancer drug response (CDR) prediction, integrating diverse information about cancer cell lines, drugs and their known responses still remains a great challenge. In this paper, we propose a graph neural network method with contrastive learning for CDR prediction. GraphCDR constructs a graph neural network based on multi-omics profiles of cancer cell lines, the chemical structure of drugs and known cancer cell line-drug responses for CDR prediction, while a contrastive learning task is presented as a regularizer within a multi-task learning paradigm to enhance the generalization ability. In the computational experiments, GraphCDR outperforms state-of-the-art methods under different experimental configurations, and the ablation study reveals the key components of GraphCDR: biological features, known cancer cell line-drug responses and contrastive learning are important for the high-accuracy CDR prediction. The experimental analyses imply the predictive power of GraphCDR and its potential value in guiding anti-cancer drug selection.

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.

How much can deep learning improve prediction of the responses to drugs in cancer cell lines?

The drug response prediction problem arises from personalized medicine and drug discovery. Deep neural networks have been applied to the multi-omics data being available for over 1000 cancer cell lines and tissues for better drug response prediction. We summarize and examine state-of-the-art deep learning methods that have been published recently. Although significant progresses have been made in deep learning approach in drug response prediction, deep learning methods show their weakness for predicting the response of a drug that does not appear in the training dataset. In particular, all the five evaluated deep learning methods performed worst than the similarity-regularized matrix factorization (SRMF) method in our drug blind test. We outline the challenges in applying deep learning approach to drug response prediction and suggest unique opportunities for deep learning integrated with established bioinformatics analyses to overcome some of these challenges.