Search

Drugs that kill tumors through multiple mechanisms have the potential for broad clinical benefits. Here, we first developed an in silico multiomics approach (BipotentR) to find cancer cell–specific regulators that simultaneously modulate tumor immunity and another oncogenic pathway and then used it to identify 38 candidate immune–metabolic regulators. We show the tumor activities of these regulators stratify patients with melanoma by their response to anti–PD-1 using machine learning and deep neural approaches, which improve the predictive power of current biomarkers. The topmost identified regulator, ESRRA, is activated in immunotherapy-resistant tumors. Its inhibition killed tumors by suppressing energy metabolism and activating two immune mechanisms: (i) cytokine induction, causing proinflammatory macrophage polarization, and (ii) antigen-presentation stimulation, recruiting CD8+ T cells into tumors. We also demonstrate a wide utility of BipotentR by applying it to angiogenesis and growth suppressor evasion pathways. BipotentR (http://bipotentr.dfci.harvard.edu/) provides a resource for evaluating patient response and discovering drug targets that act simultaneously through multiple mechanisms.Significance:BipotentR presents resources for evaluating patient response and identifying targets for drugs that can kill tumors through multiple mechanisms concurrently. Inhibition of the topmost candidate target killed tumors by suppressing energy metabolism and effects on two immune mechanisms.This article is highlighted in the In This Issue feature, p. 517

Machine learning evaluation of immune-metabolic targets in predicting patient response to anti-PD1 and anti-CTLA4 combination therapy.

Drugs that kill tumors through multiple mechanisms have the potential for broad clinical benefits. Here, we first developed an in silico multiomics approach (BipotentR) to find cancer cell–specific regulators that simultaneously modulate tumor immunity and another oncogenic pathway and then used it to identify 38 candidate immune–metabolic regulators. We show the tumor activities of these regulators stratify patients with melanoma by their response to anti–PD-1 using machine learning and deep neural approaches, which improve the predictive power of current biomarkers. The topmost identified regulator, ESRRA, is activated in immunotherapy-resistant tumors. Its inhibition killed tumors by suppressing energy metabolism and activating two immune mechanisms: (i) cytokine induction, causing proinflammatory macrophage polarization, and (ii) antigen-presentation stimulation, recruiting CD8+ T cells into tumors. We also demonstrate a wide utility of BipotentR by applying it to angiogenesis and growth suppressor evasion pathways. BipotentR (http://bipotentr.dfci.harvard.edu/) provides a resource for evaluating patient response and discovering drug targets that act simultaneously through multiple mechanisms.Significance:BipotentR presents resources for evaluating patient response and identifying targets for drugs that can kill tumors through multiple mechanisms concurrently. Inhibition of the topmost candidate target killed tumors by suppressing energy metabolism and effects on two immune mechanisms.This article is highlighted in the In This Issue feature, p. 517

We detailed a comparison of BipotentR with competing and alternative approaches in Supplementary notes 1, 2, 6, and 7. Supplementary Note 3-5 describes the in vitro findings of ESRRA, the potential clinical relevance of ESRRA across cancer-types, and the association of tumor activity of ESRRA with immune infiltration.

Cancer is a predominant disease across animals. We applied a comparative genomics approach to systematically characterize genes whose conservation levels correlate positively (PC) or negatively (NC) with cancer resistance estimates across 193 vertebrates. Pathway analysis reveals that NC genes are enriched for metabolic functions and PC genes in cell cycle regulation, DNA repair, and immune response, pointing to their corresponding roles in mediating cancer risk. We find that PC genes are less tolerant to loss-of-function (LoF) mutations, are enriched in cancer driver genes, and are associated with germline mutations that increase human cancer risk. Their relevance to cancer risk is further supported via the analysis of mouse functional genomics and cancer mortality of zoo mammals’ data. In sum, our study describes a cross-species genomic analysis pointing to candidate genes that may mediate human cancer risk.

Introduction: PARP inhibitors (PARPi) are known to have a synthetic lethal (SL) relationship with BRCA1/2 mutated cancers (ie: BRCA loss confers sensitivity to PARPi), but heterogeneity in response exists. PARPi have also found utility in BRCA-non mutated cancers, suggesting other factors may contribute to response. Recently, PARPi trials in pediatric solid tumors have had mixed results. SL interactions have been shown to exist in a network relationship, as such the aim of this study was to identify a clinically relevant SL signature (including synthetic dosage lethality and synthetic rescue) that would predict response to PARPi in pediatric cancers. Methods: Pre-treatment genomic copy number alteration, RNA sequencing, and survival data from 519 pediatric cancer patients (INFORM Consortium, 91% solid tumor) were obtained. SL partner genes for PARPi were selected if co-alteration of the drug target (PARP1/2/3) and candidate gene was 1) found to be under-represented in the patient transcriptome, 2) associated with improved survival, and 3) the genes were phylogenetically similar to PARP1/2/3. The 100 gene signature was validated on two independent pediatric cancer drug screens: in-vivo orthotopic patient-derived xenograft (OPDX) mouse models (n=24) and in-vitro PRISM cell lines (n=54). It was also validated on a cisplatin (PARP surrogate) treated pediatric osteosarcoma clinical trial (n=37). Comparisons were made to 1) a signature generated with the same pipeline using adult TCGA data, 2) a validated adult synthetic lethality pipeline (SELECT), and 3) a published 60-gene PARPi response signature (BRCAness score). Results: The pediatric signature predicted response in all drug screens evaluated. Four cell line screens: Talazoparib (AUC 0.72, p = 0.003), Rucaparib (AUC 0.69, p = 0.01), Niraparib (AUC 0.67, p = 0.03), Olaparib (AUC 0.64, p = 0.06), and two OPDX screens: Talazoparib (AUC 0.74, p = 0.03), Olaparib (AUC 0.74, p = 0.03). The pediatric signature outperformed both adult signatures (AUCs 0.34-0.61, all p’s >0.05), and the BRCAness score (AUCs 0.41-0.53, p’s >0.05) except in the Olaparib cell line (BRCAness: AUC 0.68, p = 0.01). In the pediatric osteosarcoma dataset, the pediatric signature predicted death (AUC 0.68) and two-year recurrence (AUC 0.72), outperforming the other measures (death: AUC 0.49-0.66; two-year recurrence: AUCs 0.36-0.44). Discussion: We developed a pediatric derived, clinically relevant, synthetic lethal signature that predicts response to PARPi in pan-pediatric cancer drug screens and in a cisplatin treated osteosarcoma clinical trial. Our findings support the goal of identifying precision oncology frameworks to improve the care of children with cancer. Citation Format: Matthew Nagy, Fiorella Schischlik, Kun Wang, Nishanth Ulhas Nair, E. Michael Gertz, Lipika R. Pal, Natalie Jaeger, Christopher Previti, Dina ElHarouni, Stefan M. Pfister, Eytan Ruppin. Predicting response to PARP inhibitors in pediatric cancer via synthetic lethal networks [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3118.

Background: Precision oncology is gradually advancing into mainstream clinical practice, demonstrating significant survival benefits. However, eligibility and response rates remain limited in many cases, calling for better predictive biomarkers. Methods: We present ENLIGHT, a transcriptomics-based computational approach that identifies clinically relevant genetic interactions and uses them to predict a patient’s response to a variety of therapies in multiple cancer types, importantly, without training on previous treatment response data. Consequently, in addition to its ability to predict patients' response to approved and well-studied therapies, ENLIGHT can predict the response to new treatments in early development, even before clinical data has accumulated. Accordingly, we study ENLIGHT in two translationally relevant scenarios: Personalized Oncology (PO), aimed at prioritizing approved treatments to a given patient, and Clinical Trial Design (CTD), selecting the subset of most likely responders in a patient cohort. Results: Evaluating ENLIGHT’s performance on 21 blinded clinical trial datasets spanning 11 indications and 15 different drugs in the PO setting, we show that it can effectively predict a patient’s treatment response across multiple therapies and cancer types, with an overall odds ratio of 2.59 (p=3.41e-8), and a 36% increase in response rate over the baseline (p=3.30e-13). Its prediction accuracy is better than other state-of-the-art transcriptomics-based signatures. Unlike most signatures that are prognostic or provide insights for only very few, specific treatments, ENLIGHT provides matching scores to a broad range of treatments. Quite strikingly, its performance is comparable to that of supervised predictors developed for specific indications and drugs. In combination with the IFN-γ signature, ENLIGHT achieves an odds ratio larger than 4 in predicting response to immune checkpoint therapy. In the CTD scenario, our results show that by excluding non-responders ENLIGHT can enhance clinical trial success for immunotherapies and other monoclonal antibodies, achieving > 90% of the response rate attainable under an optimal exclusion strategy. Conclusion: ENLIGHT is a powerful transcriptomics-based precision oncology pipeline developed by Pangea Biomed that broadly predicts response to both extant and novel targeted and immune therapies, going beyond context-specific biomarkers. Citation Format: Gal Dinstag, Eldad D. Shulman, Efrat Elis, Doreen S. Ben-Zvi, Omer Tirosh, Eden Maimon, Isaac Meilijson, Emmanuel Elalouf, Boris Temkin, Philipp Vitkovsky, Eyal Schiff, Danh-Tai Hoang, Sanju Sinha, Nishanth Ulhas Nair, Joo Sang-Lee, Alejandro A. Schäffer, Ze'ev Ronai, Dejan Juric, Andrea B. Apolo, William L. Dahut, Stanley Lipkowitz, Raanan Berger, Razelle Kurzrock, Antonios Papanicolau-Sengos, Fatima Karzai, Mark R. Gilbert, Kenneth Aldape, Padma S. Rajagopal, Tuvik Beker, Eytan Ruppin, Ranit Aharonov. Prediction of patient response to targeted and immunotherapies from the tumor transcriptome in a wide set of indications and clinical trials [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 957.

Malignant mesothelioma is an aggressive cancer with limited treatment options and poor prognosis. Better understanding of mesothelioma genomics and transcriptomics could advance novel therapies. We performed whole-exome and RNA-sequencing of germline and tumors of 122 patients with pleural, peritoneal, and tunica-vaginalis mesothelioma. We identify a 48 gene prognostic signature that is highly predictive of mesothelioma patient survival including CCNB1, whose expression is highly predictive of patient survival on its own. Using a synthetic-lethality (SL) based pipeline for analyzing the patients’ transcriptomic data, we identified SL-based signatures predictive of response to an anti-PD1 immune checkpoint inhibitor and combination therapies with pemetrexed. These SL-profiles successfully predict the overall patient-response observed across targeted, immuno- and chemotherapies in 11 independent mesothelioma clinical trials spanning 7 different treatments. These findings lay a basis for future studies aimed specifically at testing the ability of these SL profiles to serve as treatment biomarkers in mesothelioma.

BackgroundPrecision oncology is gradually advancing into mainstream clinical practice, demonstrating significant survival benefits. However, eligibility and response rates remain limited in many cases, calling for better predictive biomarkers.MethodsWe present ENLIGHT, a transcriptomics-based computational approach that identifies clinically relevant genetic interactions and uses them to predict a patient’s response to a variety of therapies in multiple cancer types, without training on previous treatment response data. We study ENLIGHT in two translationally oriented scenarios:Personalized Oncology (PO), aimed at prioritizing treatments for a single patient, andClinical Trial Design (CTD), selecting the most likely responders in a patient cohort.FindingsEvaluating ENLIGHT’s performance on 21 blinded clinical trial datasets in the PO setting, we show that it can effectively predict a patient’s treatment response across multiple therapies and cancer types. Its prediction accuracy is better than previously published transcriptomics-based signatures and is comparable to that of supervised predictors developed for specific indications and drugs. In combination with the IFN-γsignature, ENLIGHT achieves an odds ratio larger than 4 in predicting response to immune checkpoint therapy. In the CTD scenario, ENLIGHT can potentially enhance clinical trial success for immunotherapies and other monoclonal antibodies by excluding non-responders, while overall achieving more than 90% of the response rate attainable under an optimal exclusion strategy.ConclusionENLIGHT demonstrably enhances the ability to predict therapeutic response across multiple cancer types from the bulk tumor transcriptome.FundingThis research was supported in part by the Intramural Research Program, NIH and by the Israeli Innovation Authority.

Drugs that kill tumors through multiple mechanisms have potential for broad clinical benefits, with a reduced propensity to resistance. We developed BipotentR, a computational approach to find cancer-cell-specific regulators that simultaneously modulate tumor immunity and another oncogenic pathway. Using tumor metabolism as proof-of-principle, BipotentR identified 38 candidate immune-metabolic regulators by combining epigenomes with bulk and single-cell tumor transcriptomes from patients. Inhibition of top candidate ESRRA (Estrogen Related Receptor Alpha) killed tumors by direct effects on energy metabolism and two immune mechanisms: (i) cytokine induction, causing proinflammatory macrophage polarization (ii) antigen-presentation stimulation, recruiting CD8+T cells into tumors. ESRRA is activated in immune-suppressive and immunotherapy-resistant tumors of many types, suggesting broad clinical relevance. We also applied BipotentR to angiogenesis and growth-suppressor pathways, demonstrating a widely applicable approach to identify drug targets that act simultaneously through multiple mechanisms. BipotentR is publicly available at http://bipotentr.dfci.harvard.edu/.One-Sentence SummaryBipotentR identifies targets for bipotent anticancer drugs, as shown by the energy and immune effects of ESRRA inhibition.

Cancer occurs more frequently in men while autoimmune diseases (AIDs) occur more frequently in women. To explore whether these sex biases have a common basis, we collected 170 AID incidence studies from many countries for tissues that have both a cancer type and an AID that arise from that tissue. Analyzing a total of 182 country-specific, tissue-matched cancer-AID incidence rate sex bias data pairs, we find that the sex biases observed in the incidence of AIDs and cancers that occur in the same tissue are correlated across human tissues. Among key factors that have been previously associated with sex bias in either AID or cancer incidence, we find that the sex bias in the expression of the 37 genes encoded in the mitochondrial genome and the expression of a few immune pathways stand out as common key factors whose levels across human tissues are strongly associated with these incidence rate sex biases.

Cancer occurs more frequently in men while autoimmune diseases (AIDs) occur more frequently in women. To explore whether these sex biases have a common basis, we collected 167 AID incidence studies from many countries for tissues that have both a cancer type and an AID that arise from that tissue. Analyzing a total of 182 country-specific, tissue-matched cancer-AID incidence rate sex bias data pairs, we find that, indeed, the sex biases observed in the incidence of AIDs and cancers that occur in the same tissue are positively correlated across human tissues. The common key factor whose levels across human tissues are most strongly associated with these incidence rate sex biases is the sex bias in the expression of the 37 genes encoded in the mitochondrial genome.

SummaryTargeted therapeutics have advanced cancer treatment, but single agent activity remains limited by de novo and acquired resistance. Combining targeted drugs is broadly seen as a way to improve treatment outcome, motivating the ongoing search for efficacious combinations. To identify synergistic targeted therapy combinations and study the impact of tumor heterogeneity on combination outcome, we systematically tested over 5,000 two drug combinations at multiple doses across a collection of 81 non-small cancer cell lines. Both known and novel synergistic combinations were identified. Strikingly, very few combinations yield synergy across the majority of cell line models. Importantly, synergism mainly arises due to sensitization of single agent resistant models, rather than further sensitize already sensitive cell lines, frequently via dual targeting of a single or two highly interconnected pathways. This drug combinations resource, the largest of its kind should help delineate new synergistic regimens by facilitating the understanding of drug synergism in cancer.

The efficacy of prospective cancer treatments is routinely estimated by in vitro cell-line proliferation screens. However, it is unclear whether tumor aggressiveness and patient survival are influenced more by the proliferative or the migratory properties of cancer cells. To address this question, we experimentally measured proliferation and migration phenotypes across more than 40 breast cancer cell-lines. Based on the latter, we built and validated individual predictors of breast cancer proliferation and migration levels from the cells’ transcriptomics. We then apply these predictors to estimate the proliferation and migration levels of more than 1000 TCGA breast cancer tumors. Reassuringly, both estimates increase with tumor’s aggressiveness, as qualified by its stage, grade, and subtype. However, predicted tumor migration levels are significantly more strongly associated with patient survival than the proliferation levels. We confirmed these findings by conducting siRNA knock-down experiments on the highly migratory MDA-MB-231 cell lines and deriving gene knock-down based proliferation and migration signatures. We show that cytoskeletal drugs might be more beneficial in patients with high predicted migration levels. Taken together, these results testify to the importance of migration levels in determining patient survival.

Introduction: Most cancer deaths are caused by metastasis, yet understanding how metastatic cancers adapt from their origin tissues to their target tissues remains a fundamental scientific and clinical challenge. To date, no studies have systematically analyzed the transcriptomic similarity of metastatic cancers to their target tissues in a genome wide manner. Here, we ask if the overall gene expression of primary and metastatic tumors is closer to their tissue of origin or closer to their target tissue? Next, we aim to identify the key pathways in metastatic tumors whose gene expression becomes markedly closer to their target than primary tissue of origin. Methods: We analyzed the expression profiles of: (a) primary tumors of 9 cancer types, which have metastasized to the liver, lung, or brain (TCGA data, n = 306), (b) metastatic tumors of 12 cancer types (MET500 collection, n = 194) and (c) their origin and target normal tissue samples (GTEx, n = 5,663). We computed the similarity (Euclidean distance) between the expression profiles of the tumors (either primary or metastasis) to the mean expression of their corresponding normal tissue of origin and target tissues, termed their transcriptomic distance (TD). For each tumor sample’s expression profile (either primary tumor or metastasis), we compute the ratio of its TD to the tissue of origin over its TD to the target tissue to get its TD ratio. Results: 1) We find that while most primary tumors are more similar to the tissue of origin than to the target tissues, there is a shift in expression patterns in metastatic tumors towards their target. 2) Across cancer types that metastasize to the liver, cell cycle and growth pathways are significantly transcriptomically closer to the tissue of origin; while the expression of pathways related to more specialized liver cellular functions, such as coagulation and bile acid metabolism are becoming significantly closer to the liver. 3) We tested our key findings by analyzing a matched cohort of primary breast cancers and metastasis samples, reassuringly finding that they overall recapitulate the key results emerging from the non-matched analysis. The key pathways altered when metastasizing to four target tissues (liver, lung, ovary, skin) are adipogenesis, fatty acid metabolism, coagulation, xenobiotic metabolism, and bile acid metabolism. Conclusions: Quite surprisingly, this the first systematic analysis comparing the landscape of primary and metastatic transcriptional alterations of the same cancer type, providing a genome wide view of how cancers adapt to new environments during metastasis. The systematic identification of the specific pathways whose expression shifts towards the target tissues in metastatic tumors provides important leads to their potential targeting. Citation Format: Camilo Calvo-Alcañiz, Padma S. Rajagopal, Sanju Sinha, Fiorella Schischlik, Antonios Papanicolau-Sengos, Nishanth Ulhas Nair, Eytan Ruppin. Is the transcriptome of primary and metastatic cancers closer to their origin or target tissues [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 985.

Introduction: Despite the development of two mRNA vaccines, there is an urgent unmet need of finding new antiviral strategies. One such potential antiviral strategy is to target the synthetic lethal (SL) partners of transcriptionally altered genes in infected host cells, thereby selectively killing them to halt the infection at its heels (Mast FD, JCB, 2020). Methods: Here we conduct a first proof-of-concept SL inference approach to predict anti-SARS-CoV-2 targets in a systematic genome-wide manner. This effort capitalizes on our recently published pipeline for inferring clinically relevant SL interactions in cancer (Lee et al, Cell, 2021). Based on the latter, we comprehensively analyzed multiple in vitro and in vivo bulk and single-cell RNA-sequencing datasets of SARS-CoV-2 infection to predict candidate antiviral targets that are SL with altered host genes. Importantly, as our predictions are fine-tuned based on the analysis of patients’ data, they are more likely to be of translational value. Results: Our key results are twofold:1) The predicted SL-based targets are highly enriched for genes that are reported in four SARS-CoV-2 CRISPR-Cas9 genome-wide genetic screens to inhibit growth of infected cells. 2) A subset of top predicted 26 genes were experimentally tested in a targeted siRNA screen conducted in both infected and non-infected human Caco-2 cells. Remarkably, as expected given that these targets were predicted to be SL specific with genes upregulated in infected cells, indeed, knocking down these targets reduced viral replication and cell viability only under the infected condition without harming non-infected cells. Conclusion: In summary, this study is the first to demonstrate the potential of a synthetic lethality approach to identify viral (specifically anti-SARS-CoV-2) targets. Importantly, as both single cell and bulk transcriptomics patients’ data is considered from both infected people and controls, they are more likely to be of clinical relevance. Targeting host genes identified via an SL-based approach is probably more suitable when the infection is at the early stage and host can still tolerate the loss of infected host cells. Citation Format: Lipika R. Pal, Kuoyuan Cheng, Nishanth Ulhas Nair, Laura Martin-Sancho, Sanju Sinha, Yuan Pu, Laura Riva, Xin Yin, Fiorella Schischlik, Joo Sang Lee, Sumit Chanda, Eytan Ruppin. Identifying and testing cancer-derived synthetic-lethal anti-SARS-CoV-2 targets [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 3583.

Cancer is an evolutionarily conserved disease that occurs in a wide variety of species. We applied a comparative genomics approach to systematically characterize the genes whose conservation levels significantly correlates positively (PC) or negatively (NC) with a broad spectrum of cancer-resistance estimates, computed across almost 200 vertebrate species. PC genes are enriched in pathways relevant to tumor suppression including cell cycle, DNA repair, and immune response, while NC genes are enriched with a host of metabolic pathways. The conservation levels of the PC and NC genes in a species serve to build the first genomics-based predictor of its cancer resistance score. We find that PC genes are less tolerant to loss of function (LoF) mutations, are enriched in cancer driver genes and are associated with germline mutations that increase human cancer risk. Furthermore, their expression levels are associated with lifetime cancer risk across human tissues. Finally, their knockout in mice results in increased cancer incidence. In sum, we find that many genes associated with cancer resistance across species are implicated in human cancers, pointing to several additional candidate genes that may have a functional role in human cancer.

Tremendous progress has been made to control the COVID‐19 pandemic caused by the SARS‐CoV‐2 virus. However, effective therapeutic options are still rare. Drug repurposing and combination represent practical strategies to address this urgent unmet medical need. Viruses, including coronaviruses, are known to hijack host metabolism to facilitate viral proliferation, making targeting host metabolism a promising antiviral approach. Here, we describe an integrated analysis of 12 published in vitro and human patient gene expression datasets on SARS‐CoV‐2 infection using genome‐scale metabolic modeling (GEM), revealing complicated host metabolism reprogramming during SARS‐CoV‐2 infection. We next applied the GEM‐based metabolic transformation algorithm to predict anti‐SARS‐CoV‐2 targets that counteract the virus‐induced metabolic changes. We successfully validated these targets using published drug and genetic screen data and by performing an siRNA assay in Caco‐2 cells. Further generating and analyzing RNA‐sequencing data of remdesivir‐treated Vero E6 cell samples, we predicted metabolic targets acting in combination with remdesivir, an approved anti‐SARS‐CoV‐2 drug. Our study provides clinical data‐supported candidate anti‐SARS‐CoV‐2 targets for future evaluation, demonstrating host metabolism targeting as a promising antiviral strategy., Metabolic modeling of 12 SARS‐CoV‐2 datasets identifies novel single or combinatory antiviral targets by reverting the virus‐induced host metabolic reprogramming.

Expression pattern of synthetic lethality genes is a previously unknown factor associated with cancer risk across human tissues., Various characteristics of cancers exhibit tissue specificity, including lifetime cancer risk, onset age, and cancer driver genes. Previously, the large variation in cancer risk across human tissues was found to strongly correlate with the number of stem cell divisions and abnormal DNA methylation levels. Here, we study the role of synthetic lethality in cancer risk. Analyzing normal tissue transcriptomics data in the Genotype-Tissue Expression project, we quantify the extent of co-inactivation of cancer synthetic lethal (cSL) gene pairs and find that normal tissues with more down-regulated cSL gene pairs have lower and delayed cancer risk. Consistently, more cSL gene pairs become up-regulated in cells treated by carcinogens and throughout premalignant stages in vivo. We also show that the tissue specificity of numerous tumor suppressor genes is associated with the expression of their cSL partner genes across normal tissues. Overall, our findings support the possible role of synthetic lethality in tumorigenesis.

Novel strategies are needed to identify drug targets and treatments for the COVID-19 pandemic. The altered gene expression of virus-infected host cells provides an opportunity to specifically inhibit viral propagation via targeting the synthetic lethal (SL) partners of such altered host genes. Pursuing this antiviral strategy, here we comprehensively analyzed multiple in vitro and in vivo bulk and single-cell RNA-sequencing datasets of SARS-CoV-2 infection to predict clinically relevant candidate antiviral targets that are SL with altered host genes. The predicted SL-based targets are highly enriched for infected cell inhibiting genes reported in four SARS-CoV-2 CRISPR-Cas9 genome-wide genetic screens. Integrating our predictions with the results of these screens, we further selected a focused subset of 26 genes that we experimentally tested in a targeted siRNA screen using human Caco-2 cells. Notably, as predicted, knocking down these targets reduced viral replication and cell viability only under the infected condition without harming non-infected cells. Our results are made publicly available, to facilitate their in vivo testing and further validation.

Synthetic lethal combination induces ferroptosis., Identification of targeted therapies for TNBC is an urgent medical need. Using a drug combination screen reliant on synthetic lethal interactions, we identified clinically relevant combination therapies for different TNBC subtypes. Two drug combinations targeting the BET family were further explored. The first, targeting BET and CXCR2, is specific for mesenchymal TNBC and induces apoptosis, whereas the second, targeting BET and the proteasome, is effective for major TNBC subtypes and triggers ferroptosis. Ferroptosis was induced at low drug doses and was associated with increased cellular iron and decreased glutathione levels, concomitant with reduced levels of GPX4 and key glutathione biosynthesis genes. Further functional studies, analysis of clinical datasets and breast cancer specimens revealed a unique vulnerability of TNBC to ferroptosis inducers, enrichment of ferroptosis gene signature, and differential expression of key proteins that increase labile iron and decrease glutathione levels. This study identified potent combination therapies for TNBC and unveiled ferroptosis as a promising therapeutic strategy.

Precision oncology has made significant advances in the last few years, mainly by targeting actionable mutations in cancer driver genes. However, the proportion of patients whose tumors can be targeted therapeutically remains limited. Recent studies have begun to explore the benefit of analyzing tumor transcriptomics data to guide patient treatment, raising the need for new approaches for systematically accomplishing that. Here we show that computationally derived genetic interactions can successfully predict patient response. Assembling a broad repertoire of 32 datasets spanning more than 1,500 patients and including both tumor transcriptomics and response data, we predicted the response in 17 out of 21 targeted and 8 out of 11 checkpoint therapy datasets across 8 different cancer types with considerable accuracy, without ever training on these datasets. Analyzing the recently published multi-arm WINTHER trial, we show that the fraction of patients benefitting from transcriptomic-based treatments could potentially be markedly increased from 15% to about 85% by targeting synthetic lethal vulnerabilities in their tumors. In summary, this is the first computational approach to obtain considerable predictive performance across many different targeted and immunotherapy datasets, providing a promising new way for guiding cancer treatment based on the tumor transcriptomics of cancer patients.

Introduction Non-small-cell lung cancer (NSCLC) is among the most common cancers, accounting for 85% of all lung cancer cases. Drug combinations are known to more effective than single drugs and can delay drug resistance. NSCLC has a poor prognosis and there are no approved targeted cancer drug combinations. Here we have conducted the first large-scale screen of drug combinations in NSCLC cell lines to identify new candidate synergistic combinations and map the landscape of synthetic lethal and synergistic pathways underlying them. Results We conducted the largest of its kind in vitro drug combination screening of 81 NSCLC cell-lines for 21 x 242 cancer drug combinations at 5 different doses. Our collection of approved and investigational drugs included mainly targeted drugs as well as some chemotherapy drugs. For each drug combination, we computed the synergy score using the Bliss model. Our key findings are as follows: 1. We identify drug combinations (like CHEK and WEE1 inhibitors) that are highly synergistic in many NSCLC cell lines. 2. Combinations of targeted therapies are overall more synergistic than chemotherapeutic drug combinations. 3. Synergistic combinations occur more within pathways. 4. We see that most drug combinations do not make cell lines more sensitive than the most sensitive cell line to the corresponding single agents across all the cell lines tested. 5. We identify a new class of drug combinations - called super-sensitizers - whose combined effects are larger than the effects observed for the individual drugs composing across all of the cohort cell-lines. 6. Combinations of multi-target drugs tend to be more synergistic than single-target drugs. 7. We identify and characterize a set of context-specific drug combinations that are differentially synergistic or differentially effective (i.e. viability after treatment) between wild-type vs mutant cell lines of various cancer drivers (KRAS, TP53, STK11, PIK3CA, EGFR). 8. Studying potential mechanistic underpinnings of drug synergism in NSCLC, we find that: (a) synergistic combinations target genes that have more proximal physical interactions, (b) some synergistic combinations target synthetic lethal combinations, (c) a subset of synergistic combinations have further support by NSCLC patients survival analysis. Conclusions We performed the first large in vitro drug combination screen in NSCLC. We identify top drug combinations that are highly synergistic and effective in many cell lines and characterize key biological properties and mechanisms underlying them. The screen results are provided in a user-friendly public data resource that can help facilitate the development of targeted drug combinations in lung cancer therapy. Citation Format: Nishanth Ulhas Nair, Adam Friedman, Patricia Greninger, Avinash D. Sahu, Ellen Murchie, Giovanna Stein Crowther, Joe McClanaghan, Joo Sang Lee, Daniel A. Haber, Jeffrey Engelman, Eytan Ruppin, Cyril Benes. A large cellular screen charting the landscape of synergistic drug combinations in lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 1053.

Background Malignant mesothelioma (MM) is an aggressive cancer with limited treatment options and poor prognosis. Malignant pleural mesothelioma comprises 80% of the cases and has worse outcome than malignant peritoneal mesothelioma. An in-depth knowledge of genetic, transcriptomic and immunogenic events involved in MM is critical for successful development of prognostics and therapeutic modalities. Methods We performed whole-exome sequencing of germline and tumors of 122 patients with pleural (n=59), peritoneal (n=61) and tunica vaginalis (n=2) mesothelioma, and RNA-sequencing of 100 tumors to identify pathogenic variants, somatic mutational signatures, and prognostic gene expression signatures, predictive of patient survival and tumor response to therapies. We validated our findings using the TCGA and Bueno et al. mesothelioma datasets. Results The important findings from this study include: a) Key somatic mutational signatures are associated with DNA repair pathways and BRCA1 associated protein-1 (BAP1) is the most commonly mutated gene (~13% with germline mutation). b) We identified a set of 48 genes, a “mesothelioma prognostic signature”, whose high expression level is associated with poor survival (Cox regression, FDR < 0.1). These genes are enriched for genes related to cell cycle and DNA repair. This signature is highly predictive of patient survival in two other independent, pleural mesothelioma cohorts: TCGA (Hazard ratio (HR) = 2.6, P = 6.94e-10) and Bueno et al. mesothelioma dataset (HR = 1.49, P = 4.34e-07), after controlling for age and gender. c) Among the 48 genes, the expression of CCNB1 is highly predictive of patient survival suggesting its important role in MM, possibly via its involvement in the CDK1-CCNB1-CCNF complex (HR = 2.54, P = 1.89e-08 for TCGA; HR = 1.40, P = 1.65e-05 for Bueno et al. dataset). d) Using a synthetic lethality (SL) based precision-oncology computational framework for analyzing the patients' transcriptomic data, we were able to accurately predict response to an anti-PD1 immune checkpoint inhibitor and combination therapies with pemetrexed (chemotherapy) in mesothelioma patients. The SL profiles successfully predicted the overall patient-response observed across targeted, immuno- and chemotherapies in 11 independent mesothelioma clinical trials (Spearman's ρ = 0.64, P = 0.0348). This is the first analysis shown to successfully predict overall patient-response for various treatments within a cancer type. Conclusions By analyzing the tumor genomic and transcriptomics data of a large cohort of MM patients, we identify gene expression prognostic markers predictive of patient survival and response to therapy, both as independent signatures and via their SL interactions. These findings lay a basis for the future development of personalized therapy approaches for mesothelioma patients. Citation Format: Nishanth Ulhas Nair, Qun Jiang, Jun S. Wei, Vikram A. Misra, Betsy Morrow, Leandro C. Hermida, Joo Sang Lee, Idrees Mian, Jingli Zhang, Manjistha Sengupta, Javed Khan, Eytan Ruppin, Raffit Hassan. Genomic and transcriptomic profiling of malignant mesothelioma patients identifies gene signatures predictive of survival and response to immuno and chemotherapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 667.

Hypertension (HT) is a complex systemic disease involving transcriptional changes in multiple organs. Here we systematically investigate the pan-tissue transcriptional and genetic landscape of HT spanning dozens of tissues in hundreds of individuals. We find that in several tissues, previously identified HT-linked genes are dysregulated and the gene expression profile is predictive of HT. Importantly, many expression quantitative trait loci (eQTL) SNPs associated with the population variance of the dysregulated genes are linked with blood pressure in an independent genome-wide association study, suggesting that the functional effect of HT-associated SNPs may be mediated through tissue-specific transcriptional dysregulation. Analyses of pan-tissue transcriptional dysregulation profile, as well as eQTL SNPs underlying the dysregulated genes, reveals substantial heterogeneity among the HT patients, revealing two broad groupings – a Diffused group where several tissues exhibit HT-associated molecular alterations and a Localized group where such alterations are localized to very few tissues. These two patient subgroups differ in several clinical phenotypes including respiratory, cerebrovascular, diabetes, and heart disease. These findings suggest that the Diffused and Localized subgroups may be driven by different molecular mechanisms and have different genetic underpinning.

Most patients with advanced cancer eventually acquire resistance to targeted therapies, spurring extensive efforts to identify molecular events mediating therapy resistance. Many of these events involve synthetic rescue (SR) interactions, where the reduction in cancer cell viability caused by targeted gene inactivation is rescued by an adaptive alteration of another gene (the rescuer). Here, we perform a genome‐wide in silico prediction of SR rescuer genes by analyzing tumor transcriptomics and survival data of 10,000 TCGA cancer patients. Predicted SR interactions are validated in new experimental screens. We show that SR interactions can successfully predict cancer patients’ response and emerging resistance. Inhibiting predicted rescuer genes sensitizes resistant cancer cells to therapies synergistically, providing initial leads for developing combinatorial approaches to overcome resistance proactively. Finally, we show that the SR analysis of melanoma patients successfully identifies known mediators of resistance to immunotherapy and predicts novel rescuers.

Precision oncology has made significant advances, mainly by targeting actionable mutations in cancer driver genes. Aiming to expand treatment opportunities, recent studies have begun to explore the utility of tumor transcriptome to guide patient treatment. Here, we introduce SELECT (synthetic lethality and rescue-mediated precision oncology via the transcriptome), a precision oncology framework harnessing genetic interactions to predict patient response to cancer therapy from the tumor transcriptome. SELECT is tested on a broad collection of 35 published targeted and immunotherapy clinical trials from 10 different cancer types. It is predictive of patients' response in 80% of these clinical trials and in the recent multi-arm WINTHER trial. The predictive signatures and the code are made publicly available for academic use, laying a basis for future prospective clinical studies.

The tissue-specificity of cancer and cancer risk is a fundamental open research question. Beyond advancing our understanding of carcinogenesis, elucidating the factors underlying cancer risk may also contribute to cancer prevention. Recent studies have shown that the variation in tissue cancer risk can be explained by the number of tissue stem cell divisions occurring during the lifetime and by abnormal levels of DNA methylation. While cancer risk is likely not determined by a single factor, no other factor has been reported since to account for this fundamental variation. Here we show that cancer synthetic lethality (SL) is another strong contributor of cancer risk in human tissues. SL is a well-known type of genetic interaction where cell death occurs under the combined inactivation of two paired SL genes but not either of them alone. Targeting SLs has been recognized as a highly valuable approach for cancer treatment. We hypothesized that since down-regulated cancer SL (cSL) gene pairs reduce the viability of cancer cells, they may impede the malignant transformation of normal cells, thus modulating cancer risk. Utilizing several recently published large-scale cancer SL networks, we systematically quantified the cancer SL load (defined as the fraction of down-regulated cancer SL gene pairs) in numerous normal and cancer tissues from the TCGA and GTEx datasets. Our key findings are: 1. The cSL loads are lower in cancers vs their matched normal tissues. Furthermore, we find that the cSL load decreases progressively as cancers develop from normal tissues through multiple stages of pre-malignant lesions. These results testify that high cSL load is detrimental to cancer cells, acting as a barrier to cancer development. That is, as normal cells undergo malignant transformation, they need to reactivate at least some of the down-regulated cSL genes for the emerging cancer cells to survive. 2. In accordance with these observations, we find that cSL load in normal tissues is strongly inversely correlated with their lifetime cancer risk. 3. We find that cSL load is the first identified predictor of cancer onset time across different normal tissues - higher cSL load in the younger population is associated with later onset of cancers in that tissue. 4. cSL are important contributors of the tissue/cancer-type specificity of numerous tumor suppressor genes (like BRCA1) - that is, the activity state of cSL partners of quite a few tumor suppressor genes predicts the specific tissues in which they are known to drive cancer. Taken together, these results show that synthetic lethality load in normal tissues is a novel important biological contributor of cancer risk in humans. While synthetic lethality has been attracting tremendous attention as a way to identify cancer vulnerabilities and target them, this is the first time that its role in mediating cancer development is uncovered. Citation Format: Nishanth Ulhas Nair, Kuoyuan Cheng, Joo Sang Lee, Eytan Ruppin. Synthetic lethality across normal tissues is strongly associated with cancer risk, onset, and tumor suppressor specificity [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 36.

The efficacy of prospective cancer treatments is routinely estimated by in vitro cell-line proliferation screens. However, it is unclear whether tumor aggressiveness and patient survival are influenced more by the proliferative or the migratory properties of cancer cells. To address this question, we experimentally measured proliferation and migration phenotypes across more than 40 breast cancer cell-lines. Based on the latter, we built and validated individual predictors of breast cancer proliferation and migration levels from the cells’ transcriptomics. We then apply these predictors to estimate the proliferation and migration levels of more than 1000 TCGA breast cancer tumors. Reassuringly, both estimates increase with tumor’s aggressiveness, as qualified by its stage, grade, and subtype. However, predicted tumor migration levels are significantly more strongly associated with patient survival than the proliferation levels. We confirm these finding by conducting siRNA knock-down experiments on the highly migratory MDA-MB-231 cell lines and deriving gene knock-down based proliferation and migration signatures. We show that cytoskeletal drugs might be more beneficial in patients with high predicted migration levels. Taken together, these results testify to the importance of migration levels in determining patient survival.

The tumor suppressor Hippo pathway negatively regulates the transcriptional coactivators Yes-associated protein (YAP) and transcriptional coactivator with PDZ-binding motif (TAZ) to inhibit cell growth and control organ size, whereas activation of YAP and TAZ is implicated in tumorigenesis and cancer metastasis. Here, we report that the nonreceptor tyrosine kinase PYK2 positively regulates TAZ and YAP transcriptional activity in triple-negative breast cancer (TNBC). We found that inhibition of PYK2 expression or its kinase activity substantially affects the steady-state level of TAZ and markedly facilitates its proteasomal degradation. This effect was specific to PYK2 inhibition and was not obtained by inhibition of FAK. Destabilization of TAZ was associated with profound effect of PYK2 inhibition on cell growth at low-density concomitant with reduced expression of TAZ-target genes and induction of cell apoptosis. We further show that PYK2 enhances the tyrosine phosphorylation of both TAZ and LATS1/2 and concomitantly TAZ stability, and that PYK2 protein level correlates with the level of TAZ protein in primary breast tumors. Together these observations suggest that PYK2 is an important regulator of the Hippo pathway, and its tyrosine kinase activity has a striking effect on TAZ stabilization and activation in TNBC.

Most patients with advanced cancer eventually acquire resistance to targeted therapies, spurring extensive efforts to identify molecular events mediating therapy resistance. Many of these events involvesynthetic rescue (SR) interactions, where the reduction in cancer cell viability caused by targeted gene inactivation is rescued by an adaptive alteration of another gene (therescuer). Here we perform a genome-wide prediction of SR rescuer genes by analyzing tumor transcriptomics and survival data of 10,000 TCGA cancer patients. Predicted SR interactions are validated in new experimental screens. We show that SR interactions can successfully predict cancer patients’ response and emerging resistance. Inhibiting predicted rescuer genes sensitizes resistant cancer cells to therapies synergistically, providing initial leads for developing combinatorial approaches to overcome resistance proactively. Finally, we show that the SR analysis of melanoma patients successfully identifies known mediators of resistance to immunotherapy and predicts novel rescuers.

Idiopathic dilated cardiomyopathy (DCM) is a complex disorder with a genetic and an environmental component involving multiple genes, many of which are yet to be discovered. We integrate genetic, epigenetic, transcriptomic, phenotypic, and evolutionary features into a method – Hridaya, to infer putative functional genes underlying DCM in a genome-wide fashion, using 213 human heart genomes and transcriptomes. Many genes identified by Hridaya are experimentally shown to cause cardiac complications. We validate the top predicted genes, via five different genome-wide analyses: First, the predicted genes are associated with cardiovascular functions. Second, their knockdowns in mice induce cardiac abnormalities. Third, their inhibition by drugs cause cardiac side effects in human. Fourth, they tend to have differential exon usage between DCM and normal samples. Fifth, analyzing 213 individual genotypes, we show that regulatory polymorphisms of the predicted genes are associated with elevated risk of cardiomyopathy. The stratification of DCM patients based on cardiac expression of the functional genes reveals two subgroups differing in key cardiac phenotypes. Integrating predicted functional genes with cardiomyocyte drug treatment experiments reveals novel potential drug targets. We provide a list of investigational drugs that target the newly identified functional genes that may lead to cardiac side effects.

Motivation: We have witnessed an enormous increase in ChIP-Seq data for histone modifications in the past few years. Discovering significant patterns in these data is an important problem for understanding biological mechanisms. Results: We propose probabilistic partitioning methods to discover significant patterns in ChIP-Seq data. Our methods take into account signal magnitude, shape, strand orientation and shifts. We compare our methods with some current methods and demonstrate significant improvements, especially with sparse data. Besides pattern discovery and classification, probabilistic partitioning can serve other purposes in ChIP-Seq data analysis. Specifically, we exemplify its merits in the context of peak finding and partitioning of nucleosome positioning patterns in human promoters. Availability and implementation: The software and code are available in the supplementary material. Contact: Philipp.Bucher@isb-sib.ch Supplementary information: Supplementary Data are available at Bioinformatics online.

Joint decoding of multiple speech patterns so as to improve speech recognition performance is important, especially in the presence of noise. In this paper, we propose a Multi-Pattern Viterbi algorithm (MPVA) to jointly decode and recognize multiple speech patterns for automatic speech recognition (ASR). The MPVA is a generalization of the Viterbi Algorithm to jointly decode multiple patterns given a Hidden Markov Model (HMM). Unlike the previously proposed two stage Constrained Multi-Pattern Viterbi Algorithm (CMPVA),the MPVA is a single stage algorithm. MPVA has the advantage that it cart be extended to connected word recognition (CWR) and continuous speech recognition (CSR) problems. MPVA is shown to provide better speech recognition performance than the earlier techniques: using only two repetitions of noisy speech patterns (-5 dB SNR, 10% burst noise), the word error rate using MPVA decreased by 28.5%, when compared to using individual decoding. (C) 2010 Elsevier B.V. All rights reserved.