TUGDA: task uncertainty guided domain adaptation for robust generalization of cancer drug response prediction from in vitro to in vivo settings

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.

PANCDR: precise medicine prediction using an adversarial network for cancer drug response

Pharmacogenomics aims to provide personalized therapy to patients based on their genetic variability. However, accurate prediction of cancer drug response (CDR) is challenging due to genetic heterogeneity. Since clinical data are limited, most studies predicting drug response use preclinical data to train models. However, such models might not be generalizable to external clinical data due to differences between the preclinical and clinical datasets. In this study, a Precision Medicine Prediction using an Adversarial Network for Cancer Drug Response (PANCDR) model is proposed. PANCDR consists of two sub-models, an adversarial model and a CDR prediction model. The adversarial model reduces the gap between the preclinical and clinical datasets, while the CDR prediction model extracts features and predicts responses. PANCDR was trained using both preclinical data and unlabeled clinical data. Subsequently, it was tested on external clinical data, including The Cancer Genome Atlas and brain tumor patients. PANCDR outperformed other machine learning models in predicting external test data. Our results demonstrate the robustness of PANCDR and its potential in precision medicine by recommending patient-specific drug candidates. The PANCDR codes and data are available at https://github.com/DMCB-GIST/PANCDR.

Feature selection translates drug response predictors from cell lines to patients

Targeted therapies and chemotherapies are prevalent in cancer treatment. Identification of predictive markers to stratify cancer patients who will respond to these therapies remains challenging because patient drug response data are limited. As large amounts of drug response data have been generated by cell lines, methods to efficiently translate cell-line-trained predictors to human tumors will be useful in clinical practice. Here, we propose versatile feature selection procedures that can be combined with any classifier. For demonstration, we combined the feature selection procedures with a (linear) logit model and a (non-linear) K-nearest neighbor and trained these on cell lines to result in LogitDA and KNNDA, respectively. We show that LogitDA/KNNDA significantly outperforms existing methods, e.g., a logistic model and a deep learning method trained by thousands of genes, in prediction AUC (0.70-1.00 for seven of the ten drugs tested) and is interpretable. This may be due to the fact that sample sizes are often limited in the area of drug response prediction. We further derive a novel adjustment on the prediction cutoff for LogitDA to yield a prediction accuracy of 0.70-0.93 for seven drugs, including erlotinib and cetuximab, whose pathways relevant to anti-cancer therapies are also uncovered. These results indicate that our methods can efficiently translate cell-line-trained predictors into tumors.

Deep learning methods for drug response prediction in cancer: Predominant and emerging trends

Cancer claims millions of lives yearly worldwide. While many therapies have been made available in recent years, by in large cancer remains unsolved. Exploiting computational predictive models to study and treat cancer holds great promise in improving drug development and personalized design of treatment plans, ultimately suppressing tumors, alleviating suffering, and prolonging lives of patients. A wave of recent papers demonstrates promising results in predicting cancer response to drug treatments while utilizing deep learning methods. These papers investigate diverse data representations, neural network architectures, learning methodologies, and evaluations schemes. However, deciphering promising predominant and emerging trends is difficult due to the variety of explored methods and lack of standardized framework for comparing drug response prediction models. To obtain a comprehensive landscape of deep learning methods, we conducted an extensive search and analysis of deep learning models that predict the response to single drug treatments. A total of 61 deep learning-based models have been curated, and summary plots were generated. Based on the analysis, observable patterns and prevalence of methods have been revealed. This review allows to better understand the current state of the field and identify major challenges and promising solution paths.

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.

CREAMMIST: an integrative probabilistic database for cancer drug response prediction

Extensive in vitro cancer drug screening datasets have enabled scientists to identify biomarkers and develop machine learning models for predicting drug sensitivity. While most advancements have focused on omics profiles, cancer drug sensitivity scores precalculated by the original sources are often used as-is, without consideration for variabilities between studies. It is well-known that significant inconsistencies exist between the drug sensitivity scores across datasets due to differences in experimental setups and preprocessing methods used to obtain the sensitivity scores. As a result, many studies opt to focus only on a single dataset, leading to underutilization of available data and a limited interpretation of cancer pharmacogenomics analysis. To overcome these caveats, we have developed CREAMMIST (https://creammist.mtms.dev), an integrative database that enables users to obtain an integrative dose-response curve, to capture uncertainty (or high certainty when multiple datasets well align) across five widely used cancer cell-line drug-response datasets. We utilized the Bayesian framework to systematically integrate all available dose-response values across datasets (>14 millions dose-response data points). CREAMMIST provides easy-to-use statistics derived from the integrative dose-response curves for various downstream analyses such as identifying biomarkers, selecting drug concentrations for experiments, and training robust machine learning models.

Interpretable Machine Learning Models to Predict the Resistance of Breast Cancer Patients to Doxorubicin from Their microRNA Profiles

Doxorubicin is a common treatment for breast cancer. However, not all patients respond to this drug, which sometimes causes life-threatening side effects. Accurately anticipating doxorubicin-resistant patients would therefore permit to spare them this risk while considering alternative treatments without delay. Stratifying patients based on molecular markers in their pretreatment tumors is a promising approach to advance toward this ambitious goal, but single-gene gene markers such as HER2 expression have not shown to be sufficiently predictive. The recent availability of matched doxorubicin-response and diverse molecular profiles across breast cancer patients permits now analysis at a much larger scale. 16 machine learning algorithms and 8 molecular profiles are systematically evaluated on the same cohort of patients. Only 2 of the 128 resulting models are substantially predictive, showing that they can be easily missed by a standard-scale analysis. The best model is classification and regression tree (CART) nonlinearly combining 4 selected miRNA isoforms to predict doxorubicin response (median Matthew correlation coefficient (MCC) and area under the curve (AUC) of 0.56 and 0.80, respectively). By contrast, HER2 expression is significantly less predictive (median MCC and AUC of 0.14 and 0.57, respectively). As the predictive accuracy of this CART model increases with larger training sets, its update with future data should result in even better accuracy.