Expression Levels of Therapeutic Targets as Indicators of Sensitivity to Targeted Therapeutics

Roy, Rajarshi; Winteringham, Louise N.; Lassmann, Timo; Forrest, Alistair R.R.

doi:10.1158/1535-7163.mct-19-0273

Cited by 17 publications

(7 citation statements)

References 43 publications

(40 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The drug sensitivity data in the GDSC dataset were quantified using two continuous variables, IC50 and AUC, representing half maximal inhibitory concentration and the area under the fitted dose response curve, respectively. We converted IC50 to -log 10 IC50 and use 1-AUC to define drug sensitive as the AUC values ranges from 0 to 1 45 . We eventually used only the model that predicts 1-AUC as the final pre-trained model because it had better performance in terms of Root Mean Squared Error (RMSE) on the GDSC data, with an average RMSE of 0.096 across five test folds, comparing to mean RMSE of 0.42 achieved by the model that used − log 10 IC50 as the response variable.…”

Section: Methodsmentioning

confidence: 99%

Molecular pathways enhance drug response prediction using transfer learning from cell lines to tumors and patient-derived xenografts

Tang

Powell

Gottlieb

2022

Sci Rep

View full text Add to dashboard Cite

Computational models have been successful in predicting drug sensitivity in cancer cell line data, creating an opportunity to guide precision medicine. However, translating these models to tumors remains challenging. We propose a new transfer learning workflow that transfers drug sensitivity predicting models from large-scale cancer cell lines to both tumors and patient derived xenografts based on molecular pathways derived from genomic features. We further compute feature importance to identify pathways most important to drug response prediction. We obtained good performance on tumors (AUROC = 0.77) and patient derived xenografts from triple negative breast cancers (RMSE = 0.11). Using feature importance, we highlight the association between ER-Golgi trafficking pathway in everolimus sensitivity within breast cancer patients and the role of class II histone deacetylases and interlukine-12 in response to drugs for triple-negative breast cancer. Pathway information support transfer of drug response prediction models from cell lines to tumors and can provide biological interpretation underlying the predictions, serving as a steppingstone towards usage in clinical setting.

show abstract

Section: Methodsmentioning

confidence: 99%

Molecular pathways enhance drug response prediction using transfer learning from cell lines to tumors and patient-derived xenografts

Tang

Powell

Gottlieb

2022

Sci Rep

View full text Add to dashboard Cite

show abstract

“…Related to our findings in this context, a recent large-scale study examined the sensitivity of hundreds of cancer cell lines to hundreds of drugs, which was correlated with expression of drug targets within the cells. Here, Roy et al described inverse correlations between target expression and drug sensitivity for 8% of targets, suggesting drug efficacy may not only be determined by expression levels of the drug target, but may also depend on other factors such as genetic background and other molecules that could affect drug-target interactions, including the expression of other gene family members or interacting proteins ( 40 ).…”

Section: Discussionmentioning

confidence: 99%

SPHINX-Based Combination Therapy as a Potential Novel Treatment Strategy for Acute Myeloid Leukaemia

Wodi¹,

Belali²,

Morse³

et al. 2023

Br J Biomed Sci

View full text Add to dashboard Cite

Introduction: Dysregulated alternative splicing is a prominent feature of cancer. The inhibition and knockdown of the SR splice factor kinase SRPK1 reduces tumour growth in vivo. As a result several SPRK1 inhibitors are in development including SPHINX, a 3-(trifluoromethyl)anilide scaffold. The objective of this study was to treat two leukaemic cell lines with SPHINX in combination with the established cancer drugs azacitidine and imatinib.Materials and Methods: We selected two representative cell lines; Kasumi-1, acute myeloid leukaemia, and K562, BCR-ABL positive chronic myeloid leukaemia. Cells were treated with SPHINX concentrations up to 10μM, and in combination with azacitidine (up to 1.5 μg/ml, Kasumi-1 cells) and imatinib (up to 20 μg/ml, K562 cells). Cell viability was determined by counting the proportion of live cells and those undergoing apoptosis through the detection of activated caspase 3/7. SRPK1 was knocked down with siRNA to confirm SPHINX results.Results: The effects of SPHINX were first confirmed by observing reduced levels of phosphorylated SR proteins. SPHINX significantly reduced cell viability and increased apoptosis in Kasumi-1 cells, but less prominently in K562 cells. Knockdown of SRPK1 by RNA interference similarly reduced cell viability. Combining SPHINX with azacitidine augmented the effect of azacitidine in Kasumi-1 cells. In conclusion, SPHINX reduces cell viability and increases apoptosis in the acute myeloid leukaemia cell line Kasumi-1, but less convincingly in the chronic myeloid leukaemia cell line K562.Conclusion: We suggest that specific types of leukaemia may present an opportunity for the development of SRPK1-targeted therapies to be used in combination with established chemotherapeutic drugs.

show abstract

“…K nearest neighbor imputation was applied to impute the missing AUC values. We used the normalization method to modify the drug sensitivity data matrix of CCLs ( Roy et al, 2019 ). The drugs with

>0.1 were considered to have differential sensitivity in different clusters ( Yang et al, 2021 ).…”

Section: Methodsmentioning

confidence: 99%

Regulatory pattern of abnormal promoter CpG island methylation in the glioblastoma multiforme classification

Wang

Zhao²,

Wang³

et al. 2022

Front. Genet.

View full text Add to dashboard Cite

Glioblastoma (GBM) is characterized by extensive genetic and phenotypic heterogeneity. However, it remains unexplored primarily how CpG island methylation abnormalities in promoter mediate glioblastoma typing. First, we presented a multi-omics scale map between glioblastoma sample clusters constructed based on promoter CpG island (PCGI) methylation-driven genes, using datasets including methylation profiles, expression profiles, and single-cell sequencing data from multiple highly annotated public clinical cohorts. Second, we identified differences in the tumor microenvironment between the two glioblastoma sample clusters and resolved key signaling pathways between cell clusters at the single-cell level based on comprehensive comparative analyses to investigate the reasons for survival differences between two of these clusters. Finally, we developed a diagnostic map and a prediction model for glioblastoma, and compared theoretical differences of drug sensitivity between two glioblastoma sample clusters. In summary, this study established a classification system for dissecting promoter CpG island methylation heterogeneity in glioblastoma and provides a new perspective for the diagnosis and treatment of glioblastoma.

show abstract

Expression Levels of Therapeutic Targets as Indicators of Sensitivity to Targeted Therapeutics

Cited by 17 publications

References 43 publications

Molecular pathways enhance drug response prediction using transfer learning from cell lines to tumors and patient-derived xenografts

Molecular pathways enhance drug response prediction using transfer learning from cell lines to tumors and patient-derived xenografts

SPHINX-Based Combination Therapy as a Potential Novel Treatment Strategy for Acute Myeloid Leukaemia

Regulatory pattern of abnormal promoter CpG island methylation in the glioblastoma multiforme classification

Contact Info

Product

Resources

About