Logic models to predict continuous outputs based on binary inputs with an application to personalized cancer therapy

The Hallmarks of Predictive Oncology

SIGNIFICANCE

As the field of artificial intelligence evolves rapidly, these hallmarks are intended to capture fundamental, complementary concepts necessary for the progress and timely adoption of predictive modeling in precision oncology. Through these hallmarks, we hope to establish standards and guidelines that enable the symbiotic development of artificial intelligence and precision oncology.

Computational frameworks transform antagonism to synergy in optimizing combination therapies

While drug combinations are increasingly important in disease treatment, predicting their therapeutic interactions remains challenging. This review systematically analyzes computational methods for predicting drug combination effects through multi-omics data integration. We comprehensively assess key algorithms including DrugComboRanker and AuDNNsynergy, and evaluate integration approaches encompassing kernel regression and graph networks. The review elucidates artificial intelligence applications in predicting drug synergistic and antagonistic effects.

Considerations in the search for epistasis

Epistasis refers to changes in the effect on phenotype of a unit of genetic information, such as a single nucleotide polymorphism or a gene, dependent on the context of other genetic units. Such interactions are both biologically plausible and good candidates to explain observations which are not fully explained by an additive heritability model. However, the search for epistasis has so far largely failed to recover this missing heritability. We identify key challenges and propose that future works need to leverage idealized systems, known biology and even previously identified epistatic interactions, in order to guide the search for new interactions.

Knowledge graphs facilitate prediction of drug response for acute myeloid leukemia

Acute myeloid leukemia (AML) is a highly aggressive and heterogeneous disease, underscoring the need for improved therapeutic options and methods to optimally predict responses. With the wealth of available data resources, including clinical features, multiomics analysis, and ex vivo drug screening from AML patients, development of drug response prediction models has become feasible. Knowledge graphs (KGs) embed the relationships between different entities or features, allowing for explanation of a wide breadth of drug sensitivity and resistance mechanisms. We designed AML drug response prediction models guided by KGs. Our models included engineered features, relative gene expression between marker genes for each drug and regulators (e.g., transcription factors). We identified relative gene expression of FGD4-MIR4519, NPC2-GATA2, and BCL2-NFKB2 as predictive features for venetoclax ex vivo drug response. The KG-guided models provided high accuracy in independent test sets, overcame potential platform batch effects, and provided candidate drug sensitivity biomarkers for further validation.

Trust me if you can: a survey on reliability and interpretability of machine learning approaches for drug sensitivity prediction in cancer

With the ever-increasing number of artificial intelligence (AI) systems, mitigating risks associated with their use has become one of the most urgent scientific and societal issues. To this end, the European Union passed the EU AI Act, proposing solution strategies that can be summarized under the umbrella term trustworthiness. In anti-cancer drug sensitivity prediction, machine learning (ML) methods are developed for application in medical decision support systems, which require an extraordinary level of trustworthiness. This review offers an overview of the ML landscape of methods for anti-cancer drug sensitivity prediction, including a brief introduction to the four major ML realms (supervised, unsupervised, semi-supervised, and reinforcement learning). In particular, we address the question to what extent trustworthiness-related properties, more specifically, interpretability and reliability, have been incorporated into anti-cancer drug sensitivity prediction methods over the previous decade. In total, we analyzed 36 papers with approaches for anti-cancer drug sensitivity prediction. Our results indicate that the need for reliability has hardly been addressed so far. Interpretability, on the other hand, has often been considered for model development. However, the concept is rather used intuitively, lacking clear definitions. Thus, we propose an easily extensible taxonomy for interpretability, unifying all prevalent connotations explicitly or implicitly used within the field.

Reliable anti-cancer drug sensitivity prediction and prioritization

The application of machine learning (ML) to solve real-world problems does not only bear great potential but also high risk. One fundamental challenge in risk mitigation is to ensure the reliability of the ML predictions, i.e., the model error should be minimized, and the prediction uncertainty should be estimated. Especially for medical applications, the importance of reliable predictions can not be understated. Here, we address this challenge for anti-cancer drug sensitivity prediction and prioritization. To this end, we present a novel drug sensitivity prediction and prioritization approach guaranteeing user-specified certainty levels. The developed conformal prediction approach is applicable to classification, regression, and simultaneous regression and classification. Additionally, we propose a novel drug sensitivity measure that is based on clinically relevant drug concentrations and enables a straightforward prioritization of drugs for a given cancer sample.

Improving Anticancer Drug Selection and Prioritization via Neural Learning to Rank

Personalized cancer treatment requires a thorough understanding of complex interactions between drugs and cancer cell lines in varying genetic and molecular contexts. To address this, high-throughput screening has been used to generate large-scale drug response data, facilitating data-driven computational models. Such models can capture complex drug-cell line interactions across various contexts in a fully data-driven manner. However, accurately prioritizing the most effective drugs for each cell line still remains a significant challenge. To address this, we developed multiple neural ranking approaches that leverage large-scale drug response data across multiple cell lines from diverse cancer types. Unlike existing approaches that primarily utilize regression and classification techniques for drug response prediction, we formulated the objective of drug selection and prioritization as a drug ranking problem. In this work, we proposed multiple pairwise and listwise neural ranking methods that learn latent representations of drugs and cell lines and then use those representations to score drugs in each cell line via a learnable scoring function. Specifically, we developed neural pairwise and listwise ranking methods, Pair-PushC and List-One on top of the existing methods, pLETORg and ListNet, respectively. Additionally, we proposed a novel listwise ranking method, List-All, that focuses on all the effective drugs instead of the top effective drug, unlike List-One. We also provide an exhaustive empirical evaluation with state-of-the-art regression and ranking baselines on large-scale data sets across multiple experimental settings. Our results demonstrate that our proposed ranking methods mostly outperform the best baselines with significant improvements of as much as 25.6% in terms of selecting truly effective drugs within the top 20 predicted drugs (i.e., hit@20) across 50% test cell lines. Furthermore, our analyses suggest that the learned latent spaces from our proposed methods demonstrate informative clustering structures and capture relevant underlying biological features. Moreover, our comprehensive evaluation provides a thorough and objective comparison of the performance of different methods (including our proposed ones).

A comprehensive benchmarking of machine learning algorithms and dimensionality reduction methods for drug sensitivity prediction

A major challenge of precision oncology is the identification and prioritization of suitable treatment options based on molecular biomarkers of the considered tumor. In pursuit of this goal, large cancer cell line panels have successfully been studied to elucidate the relationship between cellular features and treatment response. Due to the high dimensionality of these datasets, machine learning (ML) is commonly used for their analysis. However, choosing a suitable algorithm and set of input features can be challenging. We performed a comprehensive benchmarking of ML methods and dimension reduction (DR) techniques for predicting drug response metrics. Using the Genomics of Drug Sensitivity in Cancer cell line panel, we trained random forests, neural networks, boosting trees and elastic nets for 179 anti-cancer compounds with feature sets derived from nine DR approaches. We compare the results regarding statistical performance, runtime and interpretability. Additionally, we provide strategies for assessing model performance compared with a simple baseline model and measuring the trade-off between models of different complexity. Lastly, we show that complex ML models benefit from using an optimized DR strategy, and that standard models-even when using considerably fewer features-can still be superior in performance.

A gray box framework that optimizes a white box logical model using a black box optimizer for simulating cellular responses to perturbations

Predicting cellular responses to perturbations requires interpretable insights into molecular regulatory dynamics to perform reliable cell fate control, despite the confounding non-linearity of the underlying interactions. There is a growing interest in developing machine learning-based perturbation response prediction models to handle the non-linearity of perturbation data, but their interpretation in terms of molecular regulatory dynamics remains a challenge. Alternatively, for meaningful biological interpretation, logical network models such as Boolean networks are widely used in systems biology to represent intracellular molecular regulation. However, determining the appropriate regulatory logic of large-scale networks remains an obstacle due to the high-dimensional and discontinuous search space. To tackle these challenges, we present a scalable derivative-free optimizer trained by meta-reinforcement learning for Boolean network models. The logical network model optimized by the trained optimizer successfully predicts anti-cancer drug responses of cancer cell lines, while simultaneously providing insight into their underlying molecular regulatory mechanisms.

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.

Multi-omic characterization of ILC and ILC-like cell lines as part of ILC cell line encyclopedia (ICLE) defines new models to study potential biomarkers and explore therapeutic opportunities

Invasive lobular carcinoma (ILC), the most common histological "special type", accounts for ∼10-15% of all BC diagnoses, is characterized by unique features such as E-cadherin loss/deficiency, lower grade, hormone receptor positivity, larger diffuse tumors, and specific metastatic patterns. Despite ILC being acknowledged as a disease with distinct biology that necessitates specialized and precision medicine treatments, the further exploration of its molecular alterations with the goal of discovering new treatments has been hindered due to the scarcity of well-characterized cell line models for studying this disease. To address this, we generated the ILC Cell Line Encyclopedia (ICLE), providing a comprehensive multi-omic characterization of ILC and ILC-like cell lines. Using consensus multi-omic subtyping, we confirmed luminal status of previously established ILC cell lines and uncovered additional ILC/ILC-like cell lines with luminal features for modeling ILC disease. Furthermore, most of these luminal ILC/ILC-like cell lines also showed RNA and copy number similarity to ILC patient tumors. Similarly, ILC/ILC-like cell lines also retained molecular alterations in key ILC genes at similar frequency to both primary and metastatic ILC tumors. Importantly, ILC/ILC-like cell lines recapitulated the CDH1 alteration landscape of ILC patient tumors including enrichment of truncating mutations in and biallelic inactivation of CDH1 gene. Using whole-genome optical mapping, we uncovered novel genomic-rearrangements including novel structural variations in CDH1 and functional gene fusions and characterized breast cancer specific patterns of chromothripsis in chromosomes 8, 11 and 17. In addition, we systematically analyzed aberrant DNAm events and integrative analysis with RNA expression revealed epigenetic activation of TFAP2B - an emerging biomarker of lobular disease that is preferentially expressed in lobular disease. Finally, towards the goal of identifying novel druggable vulnerabilities in ILC, we analyzed publicly available RNAi loss of function breast cancer cell line datasets and revealed numerous putative vulnerabilities cytoskeletal components, focal adhesion and PI3K/AKT pathway in ILC/ILC-like vs NST cell lines. In summary, we addressed the lack of suitable models to study E-cadherin deficient breast cancers by first collecting both established and putative ILC models, then characterizing them comprehensively to show their molecular similarity to patient tumors along with uncovering their novel multi-omic features as well as highlighting putative novel druggable vulnerabilities. Not only we expand the array of suitable E-cadherin deficient cell lines available for modelling human-ILC disease but also employ them for studying epigenetic activation of a putative lobular biomarker as well as identifying potential druggable vulnerabilities for this disease towards enabling precision medicine research for human-ILC.

Pan-Cancer Pharmacogenomic Analysis of Patient-Derived Tumor Cells Using Clinically Relevant Drug Exposures

As a result of tumor heterogeneity and solid cancers harboring multiple molecular defects, precision medicine platforms in oncology are most effective when both genetic and pharmacologic determinants of a tumor are evaluated. Expandable patient-derived xenograft (PDX) mouse tumor and corresponding PDX culture (PDXC) models recapitulate many of the biological and genetic characteristics of the original patient tumor, allowing for a comprehensive pharmacogenomic analysis. Here, the somatic mutations of 23 matched patient tumor and PDX samples encompassing four cancers were first evaluated using next-generation sequencing (NGS). 19 antitumor agents were evaluated across 78 patient-derived tumor cultures using clinically relevant drug exposures. A binarization threshold sensitivity classification determined in culture (PDXC) was used to identify tumors that best respond to drug in vivo (PDX). Using this sensitivity classification, logic models of DNA mutations were developed for 19 antitumor agents to predict drug response. We determined that the concordance of somatic mutations across patient and corresponding PDX samples increased as variant allele frequency increased. Notable individual PDXC responses to specific drugs, as well as lineage-specific drug responses were identified. Robust responses identified in PDXC were recapitulated in vivo in PDX-bearing mice and logic modeling determined somatic gene mutation(s) defining response to specific antitumor agents. In conclusion, combining NGS of primary patient tumors, high-throughput drug screen using clinically relevant doses, and logic modeling, can provide a platform for understanding response to therapeutic drugs targeting cancer.

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.

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.

A Boolean-based machine learning framework identifies predictive biomarkers of HSP90-targeted therapy response in prostate cancer

Precision medicine has emerged as an important paradigm in oncology, driven by the significant heterogeneity of individual patients' tumour. A key prerequisite for effective implementation of precision oncology is the development of companion biomarkers that can predict response to anti-cancer therapies and guide patient selection for clinical trials and/or treatment. However, reliable predictive biomarkers are currently lacking for many anti-cancer therapies, hampering their clinical application. Here, we developed a novel machine learning-based framework to derive predictive multi-gene biomarker panels and associated expression signatures that accurately predict cancer drug sensitivity. We demonstrated the power of the approach by applying it to identify response biomarker panels for an Hsp90-based therapy in prostate cancer, using proteomic data profiled from prostate cancer patient-derived explants. Our approach employs a rational feature section strategy to maximise model performance, and innovatively utilizes Boolean algebra methods to derive specific expression signatures of the marker proteins. Given suitable data for model training, the approach is also applicable to other cancer drug agents in different tumour settings.

RAMP: response-aware multi-task learning with contrastive regularization for cancer drug response prediction

The accurate prediction of cancer drug sensitivity according to the multiomics profiles of individual patients is crucial for precision cancer medicine. However, the development of prediction models has been challenged by the complex crosstalk of input features and the resistance-dominant drug response information contained in public databases. In this study, we propose a novel multidrug response prediction framework, response-aware multitask prediction (RAMP), via a Bayesian neural network and restrict it by soft-supervised contrastive regularization. To utilize network embedding vectors as representation learning features for heterogeneous networks, we harness response-aware negative sampling, which applies cell line-drug response information to the training of network embeddings. RAMP overcomes the prediction accuracy limitation induced by the imbalance of trained response data based on the comprehensive selection and utilization of drug response features. When trained on the Genomics of Drug Sensitivity in Cancer dataset, RAMP achieved an area under the receiver operating characteristic curve > 89%, an area under the precision-recall curve > 59% and an $\textrm{F}_1$ score > 52% and outperformed previously developed methods on both balanced and imbalanced datasets. Furthermore, RAMP predicted many missing drug responses that were not included in the public databases. Our results showed that RAMP will be suitable for the high-throughput prediction of cancer drug sensitivity and will be useful for guiding cancer drug selection processes. The Python implementation for RAMP is available at https://github.com/hvcl/RAMP.

Deep reinforcement learning for personalized treatment recommendation

In precision medicine, the ultimate goal is to recommend the most effective treatment to an individual patient based on patient-specific molecular and clinical profiles, possibly high-dimensional. To advance cancer treatment, large-scale screenings of cancer cell lines against chemical compounds have been performed to help better understand the relationship between genomic features and drug response; existing machine learning approaches use exclusively supervised learning, including penalized regression and recommender systems. However, it would be more efficient to apply reinforcement learning to sequentially learn as data accrue, including selecting the most promising therapy for a patient given individual molecular and clinical features and then collecting and learning from the corresponding data. In this article, we propose a novel personalized ranking system called Proximal Policy Optimization Ranking (PPORank), which ranks the drugs based on their predicted effects per cell line (or patient) in the framework of deep reinforcement learning (DRL). Modeled as a Markov decision process, the proposed method learns to recommend the most suitable drugs sequentially and continuously over time. As a proof-of-concept, we conduct experiments on two large-scale cancer cell line data sets in addition to simulated data. The results demonstrate that the proposed DRL-based PPORank outperforms the state-of-the-art competitors based on supervised learning. Taken together, we conclude that novel methods in the framework of DRL have great potential for precision medicine and should be further studied.

Simultaneous regression and classification for drug sensitivity prediction using an advanced random forest method

Machine learning methods trained on cancer cell line panels are intensively studied for the prediction of optimal anti-cancer therapies. While classification approaches distinguish effective from ineffective drugs, regression approaches aim to quantify the degree of drug effectiveness. However, the high specificity of most anti-cancer drugs induces a skewed distribution of drug response values in favor of the more drug-resistant cell lines, negatively affecting the classification performance (class imbalance) and regression performance (regression imbalance) for the sensitive cell lines. Here, we present a novel approach called SimultAneoUs Regression and classificatiON Random Forests (SAURON-RF) based on the idea of performing a joint regression and classification analysis. We demonstrate that SAURON-RF improves the classification and regression performance for the sensitive cell lines at the expense of a moderate loss for the resistant ones. Furthermore, our results show that simultaneous classification and regression can be superior to regression or classification alone.

Bimodal gene expression in cancer patients provides interpretable biomarkers for drug sensitivity

Identifying biomarkers predictive of cancer cell response to drug treatment constitutes one of the main challenges in precision oncology. Recent large-scale cancer pharmacogenomic studies have opened new avenues of research to develop predictive biomarkers by profiling thousands of human cancer cell lines at the molecular level and screening them with hundreds of approved drugs and experimental chemical compounds. Many studies have leveraged these data to build predictive models of response using various statistical and machine learning methods. However, a common pitfall to these methods is the lack of interpretability as to how they make predictions, hindering the clinical translation of these models. To alleviate this issue, we used the recent logic modeling approach to develop a new machine learning pipeline that explores the space of bimodally expressed genes in multiple large in vitro pharmacogenomic studies and builds multivariate, nonlinear, yet interpretable logic-based models predictive of drug response. The performance of this approach was showcased in a compendium of the three largest in vitro pharmacogenomic data sets to build robust and interpretable models for 101 drugs that span 17 drug classes with high validation rates in independent datasets. These results along with in vivo and clinical validation, support a better translation of gene expression biomarkers between model systems using bimodal gene expression.

SuperDendrix algorithm integrates genetic dependencies and genomic alterations across pathways and cancer types

Recent genome-wide CRISPR-Cas9 loss-of-function screens have identified genetic dependencies across many cancer cell lines. Associations between these dependencies and genomic alterations in the same cell lines reveal phenomena such as oncogene addiction and synthetic lethality. However, comprehensive identification of such associations is complicated by complex interactions between genes across genetically heterogeneous cancer types. We introduce and apply the algorithm SuperDendrix to CRISPR-Cas9 loss-of-function screens from 769 cancer cell lines, to identify differential dependencies across cell lines and to find associations between differential dependencies and combinations of genomic alterations and cell-type-specific markers. These associations respect the position and type of interactions within pathways: for example, we observe increased dependencies on downstream activators of pathways, such as NFE2L2, and decreased dependencies on upstream activators of pathways, such as CDK6. SuperDendrix also reveals dozens of dependencies on lineage-specific transcription factors, identifies cancer-type-specific correlations between dependencies, and enables annotation of individual mutated residues.

Using SuperDendrix, Park et al. examine associations between genetic dependencies in 769 cancer cell lines. They report 127 genetic dependencies explained by combinations of mutually exclusive somatic mutations congregating into a few oncogenic pathways across cancer subtypes. These present a small number of prominent and highly specific genetic vulnerabilities in cancer.

Graphical abstract

Implementing a Functional Precision Medicine Tumor Board for Acute Myeloid Leukemia

We generated ex vivo drug-response and multiomics profiling data for a prospective series of 252 samples from 186 patients with acute myeloid leukemia (AML). A functional precision medicine tumor board (FPMTB) integrated clinical, molecular, and functional data for application in clinical treatment decisions. Actionable drugs were found for 97% of patients with AML, and the recommendations were clinically implemented in 37 relapsed or refractory patients. We report a 59% objective response rate for the individually tailored therapies, including 13 complete responses, as well as bridging five patients with AML to allogeneic hematopoietic stem cell transplantation. Data integration across all cases enabled the identification of drug response biomarkers, such as the association of IL15 overexpression with resistance to FLT3 inhibitors. Integration of molecular profiling and large-scale drug response data across many patients will enable continuous improvement of the FPMTB recommendations, providing a paradigm for individualized implementation of functional precision cancer medicine. SIGNIFICANCE: Oncogenomics data can guide clinical treatment decisions, but often such data are neither actionable nor predictive. Functional ex vivo drug testing contributes significant additional, clinically actionable therapeutic insights for individual patients with AML. Such data can be generated in four days, enabling rapid translation through FPMTB.See related commentary by Letai, p. 290.This article is highlighted in the In This Issue feature, p. 275.

Merida: a novel boolean logic based integer linear program for personalized cancer therapy

Motivation

A major goal of personalized medicine in oncology is the optimization of treatment strategies given measurements of the genetic and molecular profiles of cancer cells. To further our knowledge on drug sensitivity, machine learning techniques are commonly applied to cancer cell line panels.

Results

We present a novel integer linear programming formulation, called MEthod for Rule Identification with multi-omics DAta (MERIDA), for predicting the drug sensitivity of cancer cells. The method represents a modified version of the LOBICO method and yields easily interpretable models amenable to a Boolean logic-based interpretation. Since the proposed altered logical rules lead to an enormous acceleration of the running times of MERIDA compared to LOBICO, we cannot only consider larger input feature sets integrated from genetic and molecular omics data but also build more comprehensive models that mirror the complexity of cancer initiation and progression. Moreover, we enable the inclusion of a priori knowledge that can either stem from biomarker databases or can also be newly acquired knowledge gathered iteratively by previous runs of MERIDA. Our results show that this approach does not only lead to an improved predictive performance but also identifies a variety of putative sensitivity and resistance biomarkers. We also compare our approach to state-of-the-art machine learning methods and demonstrate the superior performance of our method. Hence, MERIDA has great potential to deepen our understanding of the molecular mechanisms causing drug sensitivity or resistance.

Availability and implementation

The corresponding code is available on github (https://github.com/unisb-bioinf/MERIDA.git).

Supplementary information

Supplementary data are available at Bioinformatics online.

Pan-cancer Network Disorders Revealed by Overall and Local Signaling Entropy

Tumor development is a process involving loss of the differentiation phenotype and acquisition of stem-like characteristics, which is driven by intracellular rewiring of signaling network. The measurement of network reprogramming and disorder would be challenging due to the complexity and heterogeneity of tumors. Here, we proposed signaling entropy (SR) to assess the degree of tumor network disorder. We calculated SR for 33 tumor types in The Cancer Genome Atlas database based on transcriptomic and proteomic data. The SR of tumors was significantly higher than that of normal samples and was highly correlated with cell stemness, cancer type, tumor grade, and metastasis. We further demonstrated the sensitivity and accuracy of using local SR in prognosis prediction and drug response evaluation. Overall, SR could reveal cancer network disorders related to tumor malignant potency, clinical prognosis, and drug response.

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 .

Personalized logical models to investigate cancer response to BRAF treatments in melanomas and colorectal cancers

The study of response to cancer treatments has benefited greatly from the contribution of different omics data but their interpretation is sometimes difficult. Some mathematical models based on prior biological knowledge of signaling pathways facilitate this interpretation but often require fitting of their parameters using perturbation data. We propose a more qualitative mechanistic approach, based on logical formalism and on the sole mapping and interpretation of omics data, and able to recover differences in sensitivity to gene inhibition without model training. This approach is showcased by the study of BRAF inhibition in patients with melanomas and colorectal cancers who experience significant differences in sensitivity despite similar omics profiles. We first gather information from literature and build a logical model summarizing the regulatory network of the mitogen-activated protein kinase (MAPK) pathway surrounding BRAF, with factors involved in the BRAF inhibition resistance mechanisms. The relevance of this model is verified by automatically assessing that it qualitatively reproduces response or resistance behaviors identified in the literature. Data from over 100 melanoma and colorectal cancer cell lines are then used to validate the model's ability to explain differences in sensitivity. This generic model is transformed into personalized cell line-specific logical models by integrating the omics information of the cell lines as constraints of the model. The use of mutations alone allows personalized models to correlate significantly with experimental sensitivities to BRAF inhibition, both from drug and CRISPR targeting, and even better with the joint use of mutations and RNA, supporting multi-omics mechanistic models. A comparison of these untrained models with learning approaches highlights similarities in interpretation and complementarity depending on the size of the datasets. This parsimonious pipeline, which can easily be extended to other biological questions, makes it possible to explore the mechanistic causes of the response to treatment, on an individualized basis.

Identifying Drug Sensitivity Subnetworks with NETPHIX

Phenotypic heterogeneity in cancer is often caused by different patterns of genetic alterations. Understanding such phenotype-genotype relationships is fundamental for the advance of personalized medicine. We develop a computational method, named NETPHIX (NETwork-to-PHenotype association with eXclusivity) to identify subnetworks of genes whose genetic alterations are associated with drug response or other continuous cancer phenotypes. Leveraging interaction information among genes and properties of cancer mutations such as mutual exclusivity, we formulate the problem as an integer linear program and solve it optimally to obtain a subnetwork of associated genes. Applied to a large-scale drug screening dataset, NETPHIX uncovered gene modules significantly associated with drug responses. Utilizing interaction information, NETPHIX modules are functionally coherent and can thus provide important insights into drug action. In addition, we show that modules identified by NETPHIX together with their association patterns can be leveraged to suggest drug combinations.

ATEN: And/Or tree ensemble for inferring accurate Boolean network topology and dynamics

MOTIVATION

Inferring gene regulatory networks from gene expression time series data is important for gaining insights into the complex processes of cell life. A popular approach is to infer Boolean networks. However, it is still a pressing open problem to infer accurate Boolean networks from experimental data that are typically short and noisy.

RESULTS

To address the problem, we propose a Boolean network inference algorithm which is able to infer accurate Boolean network topology and dynamics from short and noisy time series data. The main idea is that, for each target gene, we use an And/Or tree ensemble algorithm to select prime implicants of which each is a conjunction of a set of input genes. The selected prime implicants are important features for predicting the states of the target gene. Using these important features we then infer the Boolean function of the target gene. Finally, the Boolean functions of all target genes are combined as a Boolean network. Using the data generated from artificial and real-world gene regulatory networks, we show that our algorithm can infer more accurate Boolean network topology and dynamics from short and noisy time series data than other algorithms. Our algorithm enables us to gain better insights into complex regulatory mechanisms of cell life.

AVAILABILITY AND IMPLEMENTATION

Package ATEN is freely available at https://github.com/ningshi/ATEN.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

iGMDR: Integrated Pharmacogenetic Resource Guide to Cancer Therapy and Research

Current pharmacogenetic studies have obtained many genetic models that can predict the therapeutic efficacy of anticancer drugs. Although some of these models are of crucial importance and have been used in clinical practice, these very valuable models have not been well adopted into cancer research to promote the development of cancer therapies due to the lack of integration and standards for the existing data of the pharmacogenetic studies. For this purpose, we built a resource investigating genetic model of drug response (iGMDR), which integrates the models from in vitro and in vivo pharmacogenetic studies with different omics data from a variety of technical systems. In this study, we introduced a standardized process for all integrations, and described how users can utilize these models to gain insights into cancer. iGMDR is freely accessible at https://igmdr.modellab.cn.

Explainable artificial intelligence models using real-world electronic health record data: a systematic scoping review

OBJECTIVE

To conduct a systematic scoping review of explainable artificial intelligence (XAI) models that use real-world electronic health record data, categorize these techniques according to different biomedical applications, identify gaps of current studies, and suggest future research directions.

MATERIALS AND METHODS

We searched MEDLINE, IEEE Xplore, and the Association for Computing Machinery (ACM) Digital Library to identify relevant papers published between January 1, 2009 and May 1, 2019. We summarized these studies based on the year of publication, prediction tasks, machine learning algorithm, dataset(s) used to build the models, the scope, category, and evaluation of the XAI methods. We further assessed the reproducibility of the studies in terms of the availability of data and code and discussed open issues and challenges.

RESULTS

Forty-two articles were included in this review. We reported the research trend and most-studied diseases. We grouped XAI methods into 5 categories: knowledge distillation and rule extraction (N = 13), intrinsically interpretable models (N = 9), data dimensionality reduction (N = 8), attention mechanism (N = 7), and feature interaction and importance (N = 5).

DISCUSSION

XAI evaluation is an open issue that requires a deeper focus in the case of medical applications. We also discuss the importance of reproducibility of research work in this field, as well as the challenges and opportunities of XAI from 2 medical professionals' point of view.

CONCLUSION

Based on our review, we found that XAI evaluation in medicine has not been adequately and formally practiced. Reproducibility remains a critical concern. Ample opportunities exist to advance XAI research in medicine.

Genetic Interactions and Tissue Specificity Modulate the Association of Mutations with Drug Response

In oncology, biomarkers are widely used to predict subgroups of patients that respond to a given drug. Although clinical decisions often rely on single gene biomarkers, machine learning approaches tend to generate complex multi-gene biomarkers that are hard to interpret. Models predicting drug response based on multiple altered genes often assume that the effects of single alterations are independent. We asked whether the association of cancer driver mutations with drug response is modulated by other driver mutations or the tissue of origin. We developed an analytic framework based on linear regression to study interactions in pharmacogenomic data from two large cancer cell line panels. Starting from a model with only covariates, we included additional variables only if they significantly improved simpler models. This allows to systematically assess interactions in small, easily interpretable models. Our results show that including mutation-mutation interactions in drug response prediction models tends to improve model performance and robustness. For example, we found that TP53 mutations decrease sensitivity to BRAF inhibitors in BRAF-mutated cell lines and patient tumors, suggesting a therapeutic benefit of combining inhibition of oncogenic BRAF with reactivation of the tumor suppressor TP53. Moreover, we identified tissue-specific mutation-drug associations and synthetic lethal triplets where the simultaneous mutation of two genes sensitizes cells to a drug. In summary, our interaction-based approach contributes to a holistic view on the determining factors of drug response.

Network-based Biased Tree Ensembles (NetBiTE) for Drug Sensitivity Prediction and Drug Sensitivity Biomarker Identification in Cancer

We present the Network-based Biased Tree Ensembles (NetBiTE) method for drug sensitivity prediction and drug sensitivity biomarker identification in cancer using a combination of prior knowledge and gene expression data. Our devised method consists of a biased tree ensemble that is built according to a probabilistic bias weight distribution. The bias weight distribution is obtained from the assignment of high weights to the drug targets and propagating the assigned weights over a protein-protein interaction network such as STRING. The propagation of weights, defines neighborhoods of influence around the drug targets and as such simulates the spread of perturbations within the cell, following drug administration. Using a synthetic dataset, we showcase how application of biased tree ensembles (BiTE) results in significant accuracy gains at a much lower computational cost compared to the unbiased random forests (RF) algorithm. We then apply NetBiTE to the Genomics of Drug Sensitivity in Cancer (GDSC) dataset and demonstrate that NetBiTE outperforms RF in predicting IC50 drug sensitivity, only for drugs that target membrane receptor pathways (MRPs): RTK, EGFR and IGFR signaling pathways. We propose based on the NetBiTE results, that for drugs that inhibit MRPs, the expression of target genes prior to drug administration is a biomarker for IC50 drug sensitivity following drug administration. We further verify and reinforce this proposition through control studies on, PI3K/MTOR signaling pathway inhibitors, a drug category that does not target MRPs, and through assignment of dummy targets to MRP inhibiting drugs and investigating the variation in NetBiTE accuracy.

Kernelized rank learning for personalized drug recommendation

Motivation

Large-scale screenings of cancer cell lines with detailed molecular profiles against libraries of pharmacological compounds are currently being performed in order to gain a better understanding of the genetic component of drug response and to enhance our ability to recommend therapies given a patient's molecular profile. These comprehensive screens differ from the clinical setting in which (i) medical records only contain the response of a patient to very few drugs, (ii) drugs are recommended by doctors based on their expert judgment and (iii) selecting the most promising therapy is often more important than accurately predicting the sensitivity to all potential drugs. Current regression models for drug sensitivity prediction fail to account for these three properties.

Results

We present a machine learning approach, named Kernelized Rank Learning (KRL), that ranks drugs based on their predicted effect per cell line (patient), circumventing the difficult problem of precisely predicting the sensitivity to the given drug. Our approach outperforms several state-of-the-art predictors in drug recommendation, particularly if the training dataset is sparse, and generalizes to patient data. Our work phrases personalized drug recommendation as a new type of machine learning problem with translational potential to the clinic.

Availability and implementation

The Python implementation of KRL and scripts for running our experiments are available at https://github.com/BorgwardtLab/Kernelized-Rank-Learning.

Supplementary information

Supplementary data are available at Bioinformatics online.

Co-occurring genomic alterations in non-small-cell lung cancer biology and therapy

The impressive clinical activity of small-molecule receptor tyrosine kinase inhibitors for oncogene-addicted subgroups of non-small-cell lung cancer (for example, those driven by activating mutations in the gene encoding epidermal growth factor receptor (EGFR) or rearrangements in the genes encoding the receptor tyrosine kinases anaplastic lymphoma kinase (ALK), ROS proto-oncogene 1 (ROS1) and rearranged during transfection (RET)) has established an oncogene-centric molecular classification paradigm in this disease. However, recent studies have revealed considerable phenotypic diversity downstream of tumour-initiating oncogenes. Co-occurring genomic alterations, particularly in tumour suppressor genes such as TP53 and LKB1 (also known as STK11), have emerged as core determinants of the molecular and clinical heterogeneity of oncogene-driven lung cancer subgroups through their effects on both tumour cell-intrinsic and non-cell-autonomous cancer hallmarks. In this Review, we discuss the impact of co-mutations on the pathogenesis, biology, microenvironmental interactions and therapeutic vulnerabilities of non-small-cell lung cancer and assess the challenges and opportunities that co-mutations present for personalized anticancer therapy, as well as the expanding field of precision immunotherapy.

Sparse discriminative latent characteristics for predicting cancer drug sensitivity from genomic features

Drug screening studies typically involve assaying the sensitivity of a range of cancer cell lines across an array of anti-cancer therapeutics. Alongside these sensitivity measurements high dimensional molecular characterizations of the cell lines are typically available, including gene expression, copy number variation and genomic mutations. We propose a sparse multitask regression model which learns discriminative latent characteristics that predict drug sensitivity and are associated with specific molecular features. We use ideas from Bayesian nonparametrics to automatically infer the appropriate number of these latent characteristics. The resulting analysis couples high predictive performance with interpretability since each latent characteristic involves a typically small set of drugs, cell lines and genomic features. Our model uncovers a number of drug-gene sensitivity associations missed by single gene analyses. We functionally validate one such novel association: that increased expression of the cell-cycle regulator C/EBPδ decreases sensitivity to the histone deacetylase (HDAC) inhibitor panobinostat.

Precision Oncology: The Road Ahead

Current efforts in precision oncology largely focus on the benefit of genomics-guided therapy. Yet, advances in sequencing techniques provide an unprecedented view of the complex genetic and nongenetic heterogeneity within individual tumors. Herein, we outline the benefits of integrating genomic and transcriptomic analyses for advanced precision oncology. We summarize relevant computational approaches to detect novel drivers and genetic vulnerabilities, suitable for therapeutic exploration. Clinically relevant platforms to functionally test predicted drugs/drug combinations for individual patients are reviewed. Finally, we highlight the technological advances in single cell analysis of tumor specimens. These may ultimately lead to the development of next-generation cancer drugs, capable of tackling the hurdles imposed by genetic and phenotypic heterogeneity on current anticancer therapies.

Precision and recall oncology: combining multiple gene mutations for improved identification of drug-sensitive tumours

Cancer drug therapies are only effective in a small proportion of patients. To make things worse, our ability to identify these responsive patients before administering a treatment is generally very limited. The recent arrival of large-scale pharmacogenomic data sets, which measure the sensitivity of molecularly profiled cancer cell lines to a panel of drugs, has boosted research on the discovery of drug sensitivity markers. However, no systematic comparison of widely-used single-gene markers with multi-gene machine-learning markers exploiting genomic data has been so far conducted. We therefore assessed the performance offered by these two types of models in discriminating between sensitive and resistant cell lines to a given drug. This was carried out for each of 127 considered drugs using genomic data characterising the cell lines. We found that the proportion of cell lines predicted to be sensitive that are actually sensitive (precision) varies strongly with the drug and type of model used. Furthermore, the proportion of sensitive cell lines that are correctly predicted as sensitive (recall) of the best single-gene marker was lower than that of the multi-gene marker in 118 of the 127 tested drugs. We conclude that single-gene markers are only able to identify those drug-sensitive cell lines with the considered actionable mutation, unlike multi-gene markers that can in principle combine multiple gene mutations to identify additional sensitive cell lines. We also found that cell line sensitivities to some drugs (e.g. Temsirolimus, 17-AAG or Methotrexate) are better predicted by these machine-learning models.

Conditional Selection of Genomic Alterations Dictates Cancer Evolution and Oncogenic Dependencies

Cancer evolves through the emergence and selection of molecular alterations. Cancer genome profiling has revealed that specific events are more or less likely to be co-selected, suggesting that the selection of one event depends on the others. However, the nature of these evolutionary dependencies and their impact remain unclear. Here, we designed SELECT, an algorithmic approach to systematically identify evolutionary dependencies from alteration patterns. By analyzing 6,456 genomes from multiple tumor types, we constructed a map of oncogenic dependencies associated with cellular pathways, transcriptional readouts, and therapeutic response. Finally, modeling of cancer evolution shows that alteration dependencies emerge only under conditional selection. These results provide a framework for the design of strategies to predict cancer progression and therapeutic response.

The Pursuit of Therapeutic Biomarkers with High-Throughput Cancer Cell Drug Screens

In the last decade we have witnessed tremendous advances in our understanding of the landscape of the molecular alterations that underpin many of the most prevalent cancers, in the use of automated high-throughput platforms for high-throughput drug screens in cancer cells, in the creation of more clinically relevant cancer cell models, and lastly in the development of more useful computational approaches in the pursuit of biomarkers of drug response. Separately, each of these improvements will undoubtedly lead to improvements in the treatment of cancer patients but to fulfill the promise of truly personalized cancer medicine, we must bring these disciplines together in a truly multidisciplinary fashion.

A Landscape of Pharmacogenomic Interactions in Cancer

Systematic studies of cancer genomes have provided unprecedented insights into the molecular nature of cancer. Using this information to guide the development and application of therapies in the clinic is challenging. Here, we report how cancer-driven alterations identified in 11,289 tumors from 29 tissues (integrating somatic mutations, copy number alterations, DNA methylation, and gene expression) can be mapped onto 1,001 molecularly annotated human cancer cell lines and correlated with sensitivity to 265 drugs. We find that cell lines faithfully recapitulate oncogenic alterations identified in tumors, find that many of these associate with drug sensitivity/resistance, and highlight the importance of tissue lineage in mediating drug response. Logic-based modeling uncovers combinations of alterations that sensitize to drugs, while machine learning demonstrates the relative importance of different data types in predicting drug response. Our analysis and datasets are rich resources to link genotypes with cellular phenotypes and to identify therapeutic options for selected cancer sub-populations.

•

We integrate heterogeneous molecular data of 11,289 tumors and 1,001 cell lines

•

We measure the response of 1,001 cancer cell lines to 265 anti-cancer drugs

•

We uncover numerous oncogenic aberrations that sensitize to an anti-cancer drug

•

Our study forms a resource to identify therapeutic options for cancer sub-populations