RESUMO
Most drugs entering clinical trials fail, often related to an incomplete understanding of the mechanisms governing drug response. Machine learning techniques hold immense promise for better drug response predictions, but most have not reached clinical practice due to their lack of interpretability and their focus on monotherapies. We address these challenges by developing DrugCell, an interpretable deep learning model of human cancer cells trained on the responses of 1,235 tumor cell lines to 684 drugs. Tumor genotypes induce states in cellular subsystems that are integrated with drug structure to predict response to therapy and, simultaneously, learn biological mechanisms underlying the drug response. DrugCell predictions are accurate in cell lines and also stratify clinical outcomes. Analysis of DrugCell mechanisms leads directly to the design of synergistic drug combinations, which we validate systematically by combinatorial CRISPR, drug-drug screening in vitro, and patient-derived xenografts. DrugCell provides a blueprint for constructing interpretable models for predictive medicine.
Assuntos
Antineoplásicos/uso terapêutico , Biologia Computacional/métodos , Neoplasias/tratamento farmacológico , Antineoplásicos/farmacologia , Linhagem Celular Tumoral , Bases de Dados Factuais , Aprendizado Profundo , Ensaios de Seleção de Medicamentos Antitumorais , Sinergismo Farmacológico , Genótipo , Humanos , Neoplasias/genética , Modelagem Computacional Específica para o PacienteRESUMO
Although cancer genomes are replete with noncoding mutations, the effects of these mutations remain poorly characterized. Here we perform an integrative analysis of 930 tumor whole genomes and matched transcriptomes, identifying a network of 193 noncoding loci in which mutations disrupt target gene expression. These 'somatic eQTLs' (expression quantitative trait loci) are frequently mutated in specific cancer tissues, and the majority can be validated in an independent cohort of 3,382 tumors. Among these, we find that the effects of noncoding mutations on DAAM1, MTG2 and HYI transcription are recapitulated in multiple cancer cell lines and that increasing DAAM1 expression leads to invasive cell migration. Collectively, the noncoding loci converge on a set of core pathways, permitting a classification of tumors into pathway-based subtypes. The somatic eQTL network is disrupted in 88% of tumors, suggesting widespread impact of noncoding mutations in cancer.
Assuntos
Genes Neoplásicos , Mutação , Neoplasias/genética , Proteínas Adaptadoras de Transdução de Sinal/genética , Aldose-Cetose Isomerases/genética , Linhagem Celular Tumoral , Regulação Neoplásica da Expressão Gênica , Redes Reguladoras de Genes , Humanos , Proteínas dos Microfilamentos , Proteínas Monoméricas de Ligação ao GTP/genética , Invasividade Neoplásica/genética , Neoplasias/metabolismo , Locos de Características Quantitativas , RNA Mensageiro/genética , RNA Mensageiro/metabolismo , RNA Neoplásico/genética , RNA Neoplásico/metabolismo , RNA não Traduzido/genética , RNA não Traduzido/metabolismo , Sequenciamento Completo do Genoma , Proteínas rho de Ligação ao GTPRESUMO
Human papillomavirus (HPV)-negative head and neck squamous cell carcinoma (HNSCC) represents a distinct classification of cancer with worse expected outcomes. Of the 11 genes recurrently mutated in HNSCC, we identify a singular and substantial survival advantage for mutations in the gene encoding Nuclear Set Domain Containing Protein 1 (NSD1), a histone methyltransferase altered in approximately 10% of patients. This effect, a 55% decrease in risk of death in NSD1-mutated versus non-mutated patients, can be validated in an independent cohort. NSD1 alterations are strongly associated with widespread genome hypomethylation in the same tumors, to a degree not observed for any other mutated gene. To address whether NSD1 plays a causal role in these associations, we use CRISPR-Cas9 to disrupt NSD1 in HNSCC cell lines and find that this leads to substantial CpG hypomethylation and sensitivity to cisplatin, a standard chemotherapy in head and neck cancer, with a 40% to 50% decrease in the IC50 value. Such results are reinforced by a survey of 1,001 cancer cell lines, in which loss-of-function NSD1 mutations have an average 23% decrease in cisplatin IC50 value compared with cell lines with wild-type NSD1Significance: This study identifies a favorable subtype of HPV-negative HNSCC linked to NSD1 mutation, hypomethylation, and cisplatin sensitivity. Mol Cancer Ther; 17(7); 1585-94. ©2018 AACR.