Search

We evaluate deconvolution methods, which infer levels of immune infiltration from bulk expression of tumor samples, through a community-wide DREAM Challenge. We assess six published and 22 community-contributed methods using in vitro and in silico transcriptional profiles of admixed cancer and healthy immune cells. Several published methods predict most cell types well, though they either were not trained to evaluate all functional CD8+ T cell states or do so with low accuracy. Several community-contributed methods address this gap, including a deep learning-based approach, whose strong performance establishes the applicability of this paradigm to deconvolution. Despite being developed largely using immune cells from healthy tissues, deconvolution methods predict levels of tumor-derived immune cells well. Our admixed and purified transcriptional profiles will be a valuable resource for developing deconvolution methods, including in response to common challenges we observe across methods, such as sensitive identification of functional CD4+ T cell states., (© 2024. The Author(s).)

Genetic alterations in cancer cells trigger oncogenic transformation, a process largely mediated by the dysregulation of kinase and transcription factor (TF) activities. While the mutational profiles of thousands of tumours have been extensively characterised, the measurements of protein activities have been technically limited until recently. We compiled public data of matched genomics and (phospho)proteomics measurements for 1,110 tumours and 77 cell lines that we used to estimate activity changes in 218 kinases and 292 TFs. Co-regulation of kinase and TF activities reflects previously known regulatory relationships and allows us to dissect genetic drivers of signalling changes in cancer. We find that loss-of-function mutations are not often associated with the dysregulation of downstream targets, suggesting frequent compensatory mechanisms. Finally, we identified the activities most differentially regulated in cancer subtypes and showed how these can be linked to differences in patient survival. Our results provide broad insights into the dysregulation of protein activities in cancer and their contribution to disease severity., (© 2023 The Authors. Published under the terms of the CC BY 4.0 license.)

The Columbia Cancer Target Discovery and Development (CTD2) Center is developing PANACEA, a resource comprising dose-responses and RNA sequencing (RNA-seq) profiles of 25 cell lines perturbed with ∼400 clinical oncology drugs, to study a tumor-specific drug mechanism of action. Here, this resource serves as the basis for a DREAM Challenge assessing the accuracy and sensitivity of computational algorithms for de novo drug polypharmacology predictions. Dose-response and perturbational profiles for 32 kinase inhibitors are provided to 21 teams who are blind to the identity of the compounds. The teams are asked to predict high-affinity binding targets of each compound among ∼1,300 targets cataloged in DrugBank. The best performing methods leverage gene expression profile similarity analysis as well as deep-learning methodologies trained on individual datasets. This study lays the foundation for future integrative analyses of pharmacogenomic data, reconciliation of polypharmacology effects in different tumor contexts, and insights into network-based assessments of drug mechanisms of action., Competing Interests: A.C. is founder, equity holder, and consultant of DarwinHealth, Inc., a company that has licensed the PANACEA database used in this manuscript from Columbia University. Columbia University is also an equity holder in DarwinHealth, Inc. J.S.-R. has received funding from GSK and Sanofi and personal fees from Travere Therapeutics. B.S. received consultation fees from Turbine, Ltd. All other authors declare no competing interests., (© 2021.)

Detailed maps of the molecular basis of the disease are powerful tools for interpreting data and building predictive models. Modularity and composability are considered necessary network features for large-scale collaborative efforts to build comprehensive molecular descriptions of disease mechanisms. An effective way to create and manage large systems is to compose multiple subsystems. Composable network components could effectively harness the contributions of many individuals and enable teams to seamlessly assemble many individual components into comprehensive maps. We examine manually built versions of the RAS-RAF-MEK-ERK cascade from the Atlas of Cancer Signalling Network, PANTHER and Reactome databases and review them in terms of their reusability and composability for assembling new disease models. We identify design principles for managing complex systems that could make it easier for investigators to share and reuse network components. We demonstrate the main challenges including incompatible levels of detail and ambiguous representation of complexes and highlight the need to address these challenges., (© The Author(s) 2021. Published by Oxford University Press.)

Pathway analysis methods are frequently applied to cancer gene expression data to identify dysregulated pathways. These methods often infer pathway activity based on the expression of genes belonging to a given pathway, even though the proteins ultimately determine the activity of a given pathway. Furthermore, the association between gene expression levels and protein activities is not well-characterized. Here, we posit that pathway-based methods are effective not because of the correlation between expression and activity of members of a given pathway, but because pathway gene sets overlap with the genes regulated by transcription factors (TFs). Thus, pathway-based methods do not inform about the activity of the pathway of interest but rather reflect changes in TF activities., (© 2020 The Authors. FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies.)

Cancer is driven by genomic alterations, but the processes causing this disease are largely performed by proteins. However, proteins are harder and more expensive to measure than genes and transcripts. To catalyze developments of methods to infer protein levels from other omics measurements, we leveraged crowdsourcing via the NCI-CPTAC DREAM proteogenomic challenge. We asked for methods to predict protein and phosphorylation levels from genomic and transcriptomic data in cancer patients. The best performance was achieved by an ensemble of models, including as predictors transcript level of the corresponding genes, interaction between genes, conservation across tumor types, and phosphosite proximity for phosphorylation prediction. Proteins from metabolic pathways and complexes were the best and worst predicted, respectively. The performance of even the best-performing model was modest, suggesting that many proteins are strongly regulated through translational control and degradation. Our results set a reference for the limitations of computational inference in proteogenomics. A record of this paper's transparent peer review process is included in the Supplemental Information., Competing Interests: Declaration of Interests The authors declare no competing interests., (Copyright © 2020 The Authors. Published by Elsevier Inc. All rights reserved.)

Knowledge about the clonal evolution of a tumor can help to interpret the function of its genetic alterations by identifying initiating events and events that contribute to the selective advantage of proliferative, metastatic, and drug-resistant subclones. Clonal evolution can be reconstructed from estimates of the relative abundance (frequency) of subclone-specific alterations in tumor biopsies, which, in turn, inform on its composition. However, estimating these frequencies is complicated by the high genetic instability that characterizes many cancers. Models for genetic instability suggest that copy number alterations (CNAs) can influence mutation-frequency estimates and thus impede efforts to reconstruct tumor phylogenies. Our analysis suggested that accurate mutation frequency estimates require accounting for CNAs-a challenging endeavour using the genetic profile of a single tumor biopsy. Instead, we propose an optimization algorithm, Chimæra, to account for the effects of CNAs using profiles of multiple biopsies per tumor. Analyses of simulated data and tumor profiles suggested that Chimæra estimates are consistently more accurate than those of previously proposed methods and resulted in improved phylogeny reconstructions and subclone characterizations. Our analyses inferred recurrent initiating mutations in hepatocellular carcinomas, resolved the clonal composition of Wilms' tumors, and characterized the acquisition of mutations in drug-resistant prostate cancers.

The prediction of transcription factor (TF) activities from the gene expression of their targets (i.e., TF regulon) is becoming a widely used approach to characterize the functional status of transcriptional regulatory circuits. Several strategies and data sets have been proposed to link the target genes likely regulated by a TF, each one providing a different level of evidence. The most established ones are (1) manually curated repositories, (2) interactions derived from ChIP-seq binding data, (3) in silico prediction of TF binding on gene promoters, and (4) reverse-engineered regulons from large gene expression data sets. However, it is not known how these different sources of regulons affect the TF activity estimations and, thereby, downstream analysis and interpretation. Here we compared the accuracy and biases of these strategies to define human TF regulons by means of their ability to predict changes in TF activities in three reference benchmark data sets. We assembled a collection of TF-target interactions for 1541 human TFs and evaluated how different molecular and regulatory properties of the TFs, such as the DNA-binding domain, specificities, or mode of interaction with the chromatin, affect the predictions of TF activity. We assessed their coverage and found little overlap on the regulons derived from each strategy and better performance by literature-curated information followed by ChIP-seq data. We provide an integrated resource of all TF-target interactions derived through these strategies, with confidence scores, as a resource for enhanced prediction of TF activities., (© 2019 Garcia-Alonso et al.; Published by Cold Spring Harbor Laboratory Press.)

The effectiveness of most cancer targeted therapies is short-lived. Tumors often develop resistance that might be overcome with drug combinations. However, the number of possible combinations is vast, necessitating data-driven approaches to find optimal patient-specific treatments. Here we report AstraZeneca's large drug combination dataset, consisting of 11,576 experiments from 910 combinations across 85 molecularly characterized cancer cell lines, and results of a DREAM Challenge to evaluate computational strategies for predicting synergistic drug pairs and biomarkers. 160 teams participated to provide a comprehensive methodological development and benchmarking. Winning methods incorporate prior knowledge of drug-target interactions. Synergy is predicted with an accuracy matching biological replicates for >60% of combinations. However, 20% of drug combinations are poorly predicted by all methods. Genomic rationale for synergy predictions are identified, including ADAM17 inhibitor antagonism when combined with PIK3CB/D inhibition contrasting to synergy when combined with other PI3K-pathway inhibitors in PIK3CA mutant cells.

Many gene fusions are reported in tumours and for most their role remains unknown. As fusions are used for diagnostic and prognostic purposes, and are targets for treatment, it is crucial to assess their function in cancer. To systematically investigate the role of fusions in tumour cell fitness, we utilized RNA-sequencing data from 1011 human cancer cell lines to functionally link 8354 fusion events with genomic data, sensitivity to >350 anti-cancer drugs and CRISPR-Cas9 loss-of-fitness effects. Established clinically-relevant fusions were identified. Overall, detection of functional fusions was rare, including those involving cancer driver genes, suggesting that many fusions are dispensable for tumour fitness. Therapeutically actionable fusions involving RAF1, BRD4 and ROS1 were verified in new histologies. In addition, recurrent YAP1-MAML2 fusions were identified as activators of Hippo-pathway signaling in multiple cancer types. Our approach discriminates functional fusions, identifying new drivers of carcinogenesis and fusions that could have clinical implications.

Acquired resistance to MEK1/2 inhibitors (MEKi) arises through amplification of BRAF V600E or KRAS G13D to reinstate ERK1/2 signalling. Here we show that BRAF V600E amplification and MEKi resistance are reversible following drug withdrawal. Cells with BRAF V600E amplification are addicted to MEKi to maintain a precise level of ERK1/2 signalling that is optimal for cell proliferation and survival, and tumour growth in vivo. Robust ERK1/2 activation following MEKi withdrawal drives a p57 KIP2 -dependent G1 cell cycle arrest and senescence or expression of NOXA and cell death, selecting against those cells with amplified BRAF V600E . p57 KIP2 expression is required for loss of BRAF V600E amplification and reversal of MEKi resistance. Thus, BRAF V600E amplification confers a selective disadvantage during drug withdrawal, validating intermittent dosing to forestall resistance. In contrast, resistance driven by KRAS G13D amplification is not reversible; rather ERK1/2 hyperactivation drives ZEB1-dependent epithelial-to-mesenchymal transition and chemoresistance, arguing strongly against the use of drug holidays in cases of KRAS G13D amplification.

Functional genomics approaches can overcome limitations-such as the lack of identification of robust targets and poor clinical efficacy-that hamper cancer drug development. Here we performed genome-scale CRISPR-Cas9 screens in 324 human cancer cell lines from 30 cancer types and developed a data-driven framework to prioritize candidates for cancer therapeutics. We integrated cell fitness effects with genomic biomarkers and target tractability for drug development to systematically prioritize new targets in defined tissues and genotypes. We verified one of our most promising dependencies, the Werner syndrome ATP-dependent helicase, as a synthetic lethal target in tumours from multiple cancer types with microsatellite instability. Our analysis provides a resource of cancer dependencies, generates a framework to prioritize cancer drug targets and suggests specific new targets. The principles described in this study can inform the initial stages of drug development by contributing to a new, diverse and more effective portfolio of cancer drug targets.

Protein signaling networks are static views of dynamic processes where proteins go through many biochemical modifications such as ubiquitination and phosphorylation to propagate signals that regulate cells and can act as feed-back systems. Understanding the precise mechanisms underlying protein interactions can elucidate how signaling and cell cycle progression occur within cells in different diseases such as cancer. Large-scale protein signaling networks contain an important number of experimentally verified protein relations but lack the capability to predict the outcomes of the system, and therefore to be trained with respect to experimental measurements. Boolean Networks (BNs) are a simple yet powerful framework to study and model the dynamics of the protein signaling networks. While many BN approaches exist to model biological systems, they focus mainly on system properties, and few exist to integrate experimental data in them. In this work, we show an application of a method conceived to integrate time series phosphoproteomic data into protein signaling networks. We use a large-scale real case study from the HPN-DREAM Breast Cancer challenge. Our efficient and parameter-free method combines logic programming and model-checking to infer a family of BNs from multiple perturbation time series data of four breast cancer cell lines given a prior protein signaling network. Because each predicted BN family is cell line specific, our method highlights commonalities and discrepancies between the four cell lines. Our models have a Root Mean Square Error (RMSE) of 0.31 with respect to the testing data, while the best performant method of this HPN-DREAM challenge had a RMSE of 0.47. To further validate our results, BNs are compared with the canonical mTOR pathway showing a comparable AUROC score (0.77) to the top performing HPN-DREAM teams. In addition, our approach can also be used as a complementary method to identify erroneous experiments. These results prove our methodology as an efficient dynamic model discovery method in multiple perturbation time course experimental data of large-scale signaling networks. The software and data are publicly available at https://github.com/misbahch6/caspo-ts., Competing Interests: The authors have declared that no competing interests exist.

Patients with seemingly the same tumour can respond very differently to treatment. There are strong, well-established effects of somatic mutations on drug efficacy, but there is at-most anecdotal evidence of a germline component to drug response. Here, we report a systematic survey of how inherited germline variants affect drug susceptibility in cancer cell lines. We develop a joint analysis approach that leverages both germline and somatic variants, before applying it to screening data from 993 cell lines and 265 drugs. Surprisingly, we find that the germline contribution to variation in drug susceptibility can be as large or larger than effects due to somatic mutations. Several of the associations identified have a direct relationship to the drug target. Finally, using 17-AAG response as an example, we show how germline effects in combination with transcriptomic data can be leveraged for improved patient stratification and to identify new markers for drug sensitivity.

Background: Genome editing by CRISPR-Cas9 technology allows large-scale screening of gene essentiality in cancer. A confounding factor when interpreting CRISPR-Cas9 screens is the high false-positive rate in detecting essential genes within copy number amplified regions of the genome. We have developed the computational tool CRISPRcleanR which is capable of identifying and correcting gene-independent responses to CRISPR-Cas9 targeting. CRISPRcleanR uses an unsupervised approach based on the segmentation of single-guide RNA fold change values across the genome, without making any assumption about the copy number status of the targeted genes., Results: Applying our method to existing and newly generated genome-wide essentiality profiles from 15 cancer cell lines, we demonstrate that CRISPRcleanR reduces false positives when calling essential genes, correcting biases within and outside of amplified regions, while maintaining true positive rates. Established cancer dependencies and essentiality signals of amplified cancer driver genes are detectable post-correction. CRISPRcleanR reports sgRNA fold changes and normalised read counts, is therefore compatible with downstream analysis tools, and works with multiple sgRNA libraries., Conclusions: CRISPRcleanR is a versatile open-source tool for the analysis of CRISPR-Cas9 knockout screens to identify essential genes.

Screening drugs on patient biopsies from solid tumours has immense potential, but is challenging due to the small amount of available material. To address this, we present here a plug-based microfluidics platform for functional screening of drug combinations. Integrated Braille valves allow changing the plug composition on demand and enable collecting >1200 data points (56 different conditions with at least 20 replicates each) per biopsy. After deriving and validating efficient and specific drug combinations for two genetically different pancreatic cancer cell lines and xenograft mouse models, we additionally screen live cells from human solid tumours with no need for ex vivo culturing steps, and obtain highly specific sensitivity profiles. The entire workflow can be completed within 48 h at assay costs of less than US$ 150 per patient. We believe this can pave the way for rapid determination of optimal personalized cancer therapies.

Cancer hallmarks are evolutionary traits required by a tumour to develop. While extensively characterised, the way these traits are achieved through the accumulation of somatic mutations in key biological pathways is not fully understood. To shed light on this subject, we characterised the landscape of pathway alterations associated with somatic mutations observed in 4,415 patients across ten cancer types, using 374 orthogonal pathway gene-sets mapped onto canonical cancer hallmarks. Towards this end, we developed SLAPenrich: a computational method based on population-level statistics, freely available as an open source R package. Assembling the identified pathway alterations into sets of hallmark signatures allowed us to connect somatic mutations to clinically interpretable cancer mechanisms. Further, we explored the heterogeneity of these signatures, in terms of ratio of altered pathways associated with each individual hallmark, assuming that this is reflective of the extent of selective advantage provided to the cancer type under consideration. Our analysis revealed the predominance of certain hallmarks in specific cancer types, thus suggesting different evolutionary trajectories across cancer lineages. Finally, although many pathway alteration enrichments are guided by somatic mutations in frequently altered high-confidence cancer genes, excluding these driver mutations preserves the hallmark heterogeneity signatures, thus the detected hallmarks' predominance across cancer types. As a consequence, we propose the hallmark signatures as a ground truth to characterise tails of infrequent genomic alterations and identify potential novel cancer driver genes and networks.

Motivation: Large pharmacogenomic screenings integrate heterogeneous cancer genomic datasets as well as anti-cancer drug responses on thousand human cancer cell lines. Mining this data to identify new therapies for cancer sub-populations would benefit from common data structures, modular computational biology tools and user-friendly interfaces., Results: We have developed GDSCTools: a software aimed at the identification of clinically relevant genomic markers of drug response. The Genomics of Drug Sensitivity in Cancer (GDSC) database (www.cancerRxgene.org) integrates heterogeneous cancer genomic datasets as well as anti-cancer drug responses on a thousand cancer cell lines. Including statistical tools (analysis of variance) and predictive methods (Elastic Net), as well as common data structures, GDSCTools allows users to reproduce published results from GDSC and to implement new analytical methods. In addition, non-GDSC data resources can also be analysed since drug responses and genomic features can be encoded as CSV files., Contact: thomas.cokelaer@pasteur.fr or saezrodriguez.rwth-aachen.de or mg12@sanger.ac.uk., Supplementary Information: Supplementary data are available at Bioinformatics online.

Transcriptional dysregulation induced by aberrant transcription factors (TF) is a key feature of cancer, but its global influence on drug sensitivity has not been examined. Here, we infer the transcriptional activity of 127 TFs through analysis of RNA-seq gene expression data newly generated for 448 cancer cell lines, combined with publicly available datasets to survey a total of 1,056 cancer cell lines and 9,250 primary tumors. Predicted TF activities are supported by their agreement with independent shRNA essentiality profiles and homozygous gene deletions, and recapitulate mutant-specific mechanisms of transcriptional dysregulation in cancer. By analyzing cell line responses to 265 compounds, we uncovered numerous TFs whose activity interacts with anticancer drugs. Importantly, combining existing pharmacogenomic markers with TF activities often improves the stratification of cell lines in response to drug treatment. Our results, which can be queried freely at dorothea.opentargets.io, offer a broad foundation for discovering opportunities to refine personalized cancer therapies. Significance: Systematic analysis of transcriptional dysregulation in cancer cell lines and patient tumor specimens offers a publicly searchable foundation to discover new opportunities to refine personalized cancer therapies. Cancer Res; 78(3); 769-80. ©2017 AACR ., (©2017 American Association for Cancer Research.)

Aberrant cell signaling can cause cancer and other diseases and is a focal point of drug research. A common approach is to infer signaling activity of pathways from gene expression. However, mapping gene expression to pathway components disregards the effect of post-translational modifications, and downstream signatures represent very specific experimental conditions. Here we present PROGENy, a method that overcomes both limitations by leveraging a large compendium of publicly available perturbation experiments to yield a common core of Pathway RespOnsive GENes. Unlike pathway mapping methods, PROGENy can (i) recover the effect of known driver mutations, (ii) provide or improve strong markers for drug indications, and (iii) distinguish between oncogenic and tumor suppressor pathways for patient survival. Collectively, these results show that PROGENy accurately infers pathway activity from gene expression in a wide range of conditions.

Proteome balance is safeguarded by the proteostasis network (PN), an intricately regulated network of conserved processes that evolved to maintain native function of the diverse ensemble of protein species, ensuring cellular and organismal health. Proteostasis imbalances and collapse are implicated in a spectrum of human diseases, from neurodegeneration to cancer. The characteristics of PN disease alterations however have not been assessed in a systematic way. Since the chaperome is among the central components of the PN, we focused on the chaperome in our study by utilizing a curated functional ontology of the human chaperome that we connect in a high-confidence physical protein-protein interaction network. Challenged by the lack of a systems-level understanding of proteostasis alterations in the heterogeneous spectrum of human cancers, we assessed gene expression across more than 10,000 patient biopsies covering 22 solid cancers. We derived a novel customized Meta-PCA dimension reduction approach yielding M-scores as quantitative indicators of disease expression changes to condense the complexity of cancer transcriptomics datasets into quantitative functional network topographies. We confirm upregulation of the HSP90 family and also highlight HSP60s, Prefoldins, HSP100s, ER- and mitochondria-specific chaperones as pan-cancer enriched. Our analysis also reveals a surprisingly consistent strong downregulation of small heat shock proteins (sHSPs) and we stratify two cancer groups based on the preferential upregulation of ATP-dependent chaperones. Strikingly, our analyses highlight similarities between stem cell and cancer proteostasis, and diametrically opposed chaperome deregulation between cancers and neurodegenerative diseases. We developed a web-based Proteostasis Profiler tool (Pro2) enabling intuitive analysis and visual exploration of proteostasis disease alterations using gene expression data. Our study showcases a comprehensive profiling of chaperome shifts in human cancers and sets the stage for a systematic global analysis of PN alterations across the human diseasome towards novel hypotheses for therapeutic network re-adjustment in proteostasis disorders.

Deregulated signal transduction and energy metabolism are hallmarks of cancer and both play a fundamental role in tumorigenesis. While it is increasingly recognised that signalling and metabolism are highly interconnected, the underpinning mechanisms of their co-regulation are still largely unknown. Here we designed and acquired proteomics, phosphoproteomics, and metabolomics experiments in fumarate hydratase (FH) deficient cells and developed a computational modelling approach to identify putative regulatory phosphorylation-sites of metabolic enzymes. We identified previously reported functionally relevant phosphosites and potentially novel regulatory residues in enzymes of the central carbon metabolism. In particular, we showed that pyruvate dehydrogenase (PDHA1) enzymatic activity is inhibited by increased phosphorylation in FH-deficient cells, restricting carbon entry from glucose to the tricarboxylic acid cycle. Moreover, we confirmed PDHA1 phosphorylation in human FH-deficient tumours. Our work provides a novel approach to investigate how post-translational modifications of enzymes regulate metabolism and could have important implications for understanding the metabolic transformation of FH-deficient cancers with potential clinical applications., (Copyright © 2017 The Authors. Published by Elsevier Inc. All rights reserved.)

Cellular signaling, predominantly mediated by phosphorylation through protein kinases, is found to be deregulated in most cancers. Accordingly, protein kinases have been subject to intense investigations in cancer research, to understand their role in oncogenesis and to discover new therapeutic targets. Despite great advances, an understanding of kinase dysfunction in cancer is far from complete.A powerful tool to investigate phosphorylation is mass-spectrometry (MS)-based phosphoproteomics, which enables the identification of thousands of phosphorylated peptides in a single experiment. Since every phosphorylation event results from the activity of a protein kinase, high-coverage phosphoproteomics data should indirectly contain comprehensive information about the activity of protein kinases.In this chapter, we discuss the use of computational methods to predict kinase activity scores from MS-based phosphoproteomics data. We start with a short explanation of the fundamental features of the phosphoproteomics data acquisition process from the perspective of the computational analysis. Next, we briefly review the existing databases with experimentally verified kinase-substrate relationships and present a set of bioinformatic tools to discover novel kinase targets. We then introduce different methods to infer kinase activities from phosphoproteomics data and these kinase-substrate relationships. We illustrate their application with a detailed protocol of one of the methods, KSEA (Kinase Substrate Enrichment Analysis). This method is implemented in Python within the framework of the open-source Kinase Activity Toolbox (kinact), which is freely available at http://github.com/saezlab/kinact/ .

Copy-number variations (CNVs) are ubiquitous in cancer and often act as driver events, but the effects of CNVs on the proteome of tumors are poorly understood. Here, we analyze recently published genomics, transcriptomics, and proteomics datasets made available by CPTAC and TCGA consortia on 282 breast, ovarian, and colorectal tumor samples to investigate the impact of CNVs in the proteomes of these cells. We found that CNVs are buffered by post-transcriptional regulation in 23%-33% of proteins that are significantly enriched in protein complex members. Our analyses show that complex subunits are highly co-regulated, and some act as rate-limiting steps of complex assembly, as their depletion induces decreased abundance of other complex members. We identified 48 such rate-limiting interactions and experimentally confirmed our predictions on the interactions of AP3B1 with AP3M1 and GTF2E2 with GTF2E1. This study highlights the importance of post-transcriptional mechanisms in cancer that allow cells to cope with their altered genomes., (Copyright © 2017 The Authors. Published by Elsevier Inc. All rights reserved.)

Combinations of therapies are being actively pursued to expand therapeutic options and deal with cancer's pervasive resistance to treatment. Research efforts to discover effective combination treatments have focused on drugs targeting intracellular processes of the cancer cells and in particular on small molecules that target aberrant kinases. Accordingly, most of the computational methods used to study, predict, and develop drug combinations concentrate on these modes of action and signaling processes within the cancer cell. This focus on the cancer cell overlooks significant opportunities to tackle other components of tumor biology that may offer greater potential for improving patient survival. Many alternative strategies have been developed to combat cancer; for example, targeting different cancer cellular processes such as epigenetic control; modulating stromal cells that interact with the tumor; strengthening physical barriers that confine tumor growth; boosting the immune system to attack tumor cells; and even regulating the microbiome to support antitumor responses. We suggest that to fully exploit these treatment modalities using effective drug combinations it is necessary to develop multiscale computational approaches that take into account the full complexity underlying the biology of a tumor, its microenvironment, and a patient's response to the drugs. In this Opinion article, we discuss preliminary work in this area and the needs-in terms of both computational and data requirements-that will truly empower such combinations.

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., (Copyright © 2016 The Author(s). Published by Elsevier Inc. All rights reserved.)

It remains unclear whether causal, rather than merely correlational, relationships in molecular networks can be inferred in complex biological settings. Here we describe the HPN-DREAM network inference challenge, which focused on learning causal influences in signaling networks. We used phosphoprotein data from cancer cell lines as well as in silico data from a nonlinear dynamical model. Using the phosphoprotein data, we scored more than 2,000 networks submitted by challenge participants. The networks spanned 32 biological contexts and were scored in terms of causal validity with respect to unseen interventional data. A number of approaches were effective, and incorporating known biology was generally advantageous. Additional sub-challenges considered time-course prediction and visualization. Our results suggest that learning causal relationships may be feasible in complex settings such as disease states. Furthermore, our scoring approach provides a practical way to empirically assess inferred molecular networks in a causal sense.

We present a novel strategy to identify drug-repositioning opportunities. The starting point of our method is the generation of a signature summarising the consensual transcriptional response of multiple human cell lines to a compound of interest (namely the seed compound). This signature can be derived from data in existing databases, such as the connectivity-map, and it is used at first instance to query a network interlinking all the connectivity-map compounds, based on the similarity of their transcriptional responses. This provides a drug neighbourhood, composed of compounds predicted to share some effects with the seed one. The original signature is then refined by systematically reducing its overlap with the transcriptional responses induced by drugs in this neighbourhood that are known to share a secondary effect with the seed compound. Finally, the drug network is queried again with the resulting refined signatures and the whole process is carried on for a number of iterations. Drugs in the final refined neighbourhood are then predicted to exert the principal mode of action of the seed compound. We illustrate our approach using paclitaxel (a microtubule stabilising agent) as seed compound. Our method predicts that glipizide and splitomicin perturb microtubule function in human cells: a result that could not be obtained through standard signature matching methods. In agreement, we find that glipizide and splitomicin reduce interphase microtubule growth rates and transiently increase the percentage of mitotic cells-consistent with our prediction. Finally, we validated the refined signatures of paclitaxel response by mining a large drug screening dataset, showing that human cancer cell lines whose basal transcriptional profile is anti-correlated to them are significantly more sensitive to paclitaxel and docetaxel.

Cell signaling pathways control cells' responses to their environment through an intricate network of proteins and small molecules partitioned by intracellular structures, such as the cytoskeleton and nucleus. Our understanding of these pathways has been revised recently with the advent of more advanced experimental techniques; no longer are signaling pathways viewed as linear cascades of information flowing from membrane-bound receptors to the nucleus. Instead, such pathways must be understood in the context of networks, and studying such networks requires an integration of computational and experimental approaches. This understanding is becoming more important in designing novel therapies for diseases such as cancer. Using the MAPK (mitogen-activated protein kinase) and PI3K (class I phosphoinositide-3' kinase) pathways as case studies of cellular signaling, we give an overview of these pathways and their functions. We then describe, using a number of case studies, how computational modeling has aided in understanding these pathways' deregulation in cancer, and how such understanding can be used to optimally tailor current therapies or help design new therapies against cancer.

Predicting the best treatment strategy from genomic information is a core goal of precision medicine. Here we focus on predicting drug response based on a cohort of genomic, epigenomic and proteomic profiling data sets measured in human breast cancer cell lines. Through a collaborative effort between the National Cancer Institute (NCI) and the Dialogue on Reverse Engineering Assessment and Methods (DREAM) project, we analyzed a total of 44 drug sensitivity prediction algorithms. The top-performing approaches modeled nonlinear relationships and incorporated biological pathway information. We found that gene expression microarrays consistently provided the best predictive power of the individual profiling data sets; however, performance was increased by including multiple, independent data sets. We discuss the innovations underlying the top-performing methodology, Bayesian multitask MKL, and we provide detailed descriptions of all methods. This study establishes benchmarks for drug sensitivity prediction and identifies approaches that can be leveraged for the development of new methods.

Predicting the response of a specific cancer to a therapy is a major goal in modern oncology that should ultimately lead to a personalised treatment. High-throughput screenings of potentially active compounds against a panel of genomically heterogeneous cancer cell lines have unveiled multiple relationships between genomic alterations and drug responses. Various computational approaches have been proposed to predict sensitivity based on genomic features, while others have used the chemical properties of the drugs to ascertain their effect. In an effort to integrate these complementary approaches, we developed machine learning models to predict the response of cancer cell lines to drug treatment, quantified through IC₅₀ values, based on both the genomic features of the cell lines and the chemical properties of the considered drugs. Models predicted IC₅₀ values in a 8-fold cross-validation and an independent blind test with coefficient of determination R² of 0.72 and 0.64 respectively. Furthermore, models were able to predict with comparable accuracy (R² of 0.61) IC50s of cell lines from a tissue not used in the training stage. Our in silico models can be used to optimise the experimental design of drug-cell screenings by estimating a large proportion of missing IC₅₀ values rather than experimentally measuring them. The implications of our results go beyond virtual drug screening design: potentially thousands of drugs could be probed in silico to systematically test their potential efficacy as anti-tumour agents based on their structure, thus providing a computational framework to identify new drug repositioning opportunities as well as ultimately be useful for personalized medicine by linking the genomic traits of patients to drug sensitivity.

Clinical responses to anticancer therapies are often restricted to a subset of patients. In some cases, mutated cancer genes are potent biomarkers for responses to targeted agents. Here, to uncover new biomarkers of sensitivity and resistance to cancer therapeutics, we screened a panel of several hundred cancer cell lines--which represent much of the tissue-type and genetic diversity of human cancers--with 130 drugs under clinical and preclinical investigation. In aggregate, we found that mutated cancer genes were associated with cellular response to most currently available cancer drugs. Classic oncogene addiction paradigms were modified by additional tissue-specific or expression biomarkers, and some frequently mutated genes were associated with sensitivity to a broad range of therapeutic agents. Unexpected relationships were revealed, including the marked sensitivity of Ewing's sarcoma cells harbouring the EWS (also known as EWSR1)-FLI1 gene translocation to poly(ADP-ribose) polymerase (PARP) inhibitors. By linking drug activity to the functional complexity of cancer genomes, systematic pharmacogenomic profiling in cancer cell lines provides a powerful biomarker discovery platform to guide rational cancer therapeutic strategies.