Gil Speyer

KNOWLEDGE-ASSISTED APPROACH TO IDENTIFY PATHWAYS WITH DIFFERENTIAL DEPENDENCIES *

We have previously developed a statistical method to identify gene sets enriched with condition-specific genetic dependencies. The method constructs gene dependency networks from bootstrapped samples in one condition and computes the divergence between distributions of network likelihood scores from different conditions. It was shown to be capable of sensitive and specific identification of pathways with phenotype-specific dysregulation, i.e., rewiring of dependencies between genes in different conditions. We now present an extension of the method by incorporating prior knowledge into the inference of networks. The degree of prior knowledge incorporation has substantial effect on the sensitivity of the method, as the data is the source of condition specificity while prior knowledge incorporation can provide additional support for dependencies that are only partially supported by the data. Use of prior knowledge also significantly improved the interpretability of the results. Further analysis of topological characteristics of gene differential dependency networks provides a new approach to identify genes that could play important roles in biological signaling in a specific condition, hence, promising targets customized to a specific condition. Through analysis of TCGA glioblastoma multiforme data, we demonstrate the method can identify not only potentially promising targets but also underlying biology for new targets.

Sex-specific gene and pathway modeling of inherited glioma risk

Background To date, genome-wide association studies (GWAS) have identified 25 risk variants for glioma, explaining 30% of heritable risk. Most histologies occur with significantly higher incidence in males, and this difference is not explained by currently known risk factors. A previous GWAS identified sex-specific glioma risk variants, and this analysis aims to further elucidate risk variation by sex using gene- and pathway-based approaches. Methods Results from the Glioma International Case-Control Study were used as a testing set, and results from 3 GWAS were combined via meta-analysis and used as a validation set. Using summary statistics for nominally significant autosomal SNPs (P < 0.01 in a previous meta-analysis) and nominally significant X-chromosome SNPs (P < 0.01), 3 algorithms (Pascal, BimBam, and GATES) were used to generate gene scores, and Pascal was used to generate pathway scores. Results were considered statistically significant in the discovery set when P < 3.3 × 10-6 and in the validation set when P < 0.001 in 2 of 3 algorithms. Results Twenty-five genes within 5 regions and 19 genes within 6 regions reached statistical significance in at least 2 of 3 algorithms in males and females, respectively. EGFR was significantly associated with all glioma and glioblastoma in males only and a female-specific association in TERT, all of which remained nominally significant after conditioning on known risk loci. There were nominal associations with the BioCarta telomeres pathway in both males and females. Conclusions These results provide additional evidence that there may be differences by sex in genetic risk for glioma. Additional analyses may further elucidate the biological processes through which this risk is conferred.

Abstract 1142: Novel target discovery for glioblastoma using chemical biology fingerprinting

The most common adult brain tumor is Glioblastoma Multiforme (GBM), an extremely aggressive cancer with only scant treatment options. Even with standard of care most patients present with a recurrence and the median survival is only circa 15 months. The need, therefore, for new therapeutic targets and treatment options is pressing. Here we describe here a multipronged approach to identifying said targets. We present an established methodology for the isolation and culture of patient derived GBM samples that retain the “stem-like” fraction thought to underlie resistance and recurrence. Furthermore we show genomically that these samples represent specific subtypes of the disease yet still form distinct groups in unbiased clustering analysis. Thus we have multiple representative patient derived cultures that are suitable for our drug discovery and chemical biology analyses. Using a process we term Chemical Biology Fingerprinting (CBF) we utilize small focused, and clinically relevant, chemical collections in order to identify patterns of chemovulnerabilities across multiple samples. This allows an unbiased yet cancer relevant sub-stratification and the identification of agents, and therefore targets, which may be relevant for GBM patient subtypes. Indeed our use of the highly annotated NCI CTD2 Informer Set of chemicals allows ready drug-to-target mapping and facilitates data sharing across the CTD2 network. Moreover, already defined subgroups can be clustered to find agents, or groups of agents, that show selective activity against traditional classifications (e.g. proneural, mesenchymal etc.). Finally our strategy is permissive for the identification of “exceptional responders”. That is, individual patient samples that respond to a specific drug whilst most samples are refractory. In sum we demonstrate generation of patient derived models and identify specific, and novel, drugs that may be relevant for specific GBM subtypes. Supported by NIH U01CA168397 Citation Format: Darren Finlay, Pedro Aza-Blanc, Harshil Dhruv, Alexey Eroshkin, Craig Hauser, Jeff Kiefer, Seungchan Kim, Tao Long, Robert G. Oshima, Sen Peng, Gil Speyer, Michael Berens, Kristiina Vuori. Novel target discovery for glioblastoma using chemical biology fingerprinting [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 1142. doi:10.1158/1538-7445.AM2017-1142

Abstract A32: Comprehensive analysis of molecular pathogenesis of multiple myeloma by genetic ancestry

Introduction: Multiple Myeloma (MM) is a complex malignancy of plasma cells triggered by immunoglobulin gene rearrangements and well-described hyperdyploidy, accounting for slightly more than 10% of all hematologic cancers. MM is the most common hematologic malignancy in African American population with conspicuous racial disparities in both, mortality rates and incidence in cancer compared to European American. This observation evolved into a central hypothesis that MM has distinct biological differences across different ethnicities with yet unidentified race specific markers of tumor heterogeneity. Clear understanding of these molecular differences among ethnic minorities with MM will fulfill a major unmet medical need and eliminate racial disparity. Methods: We acquired data from the Multiple Myeloma Research Foundation (MMRF) initiated comprehensive longitudinal study (CoMMpass) with an overall goal to profile 1,000 multiple myeloma patients at diagnosis, with multiple follow-up points throughout the course of the disease. To generate population subgroups based on genetic ancestry, we used a population stratification principle component analysis (PCA) and STRUCTURE to stratify myeloma patients by Ancestry Informative Markers. These well-established methods have allowed us to avoid confounders associated with self-reporting, and thus stratify the myeloma samples by genetic ancestry mapped along with anchor populations developed by 1000 genome project. We then assessed mutational frequencies as a function of PCA for each ancestry group using complex bioinformatics algorithms. Results: We confirmed known commonly mutated genes in MM including KRAS, NRAS, FAM46C, and DIS3. Among the most striking and novel observations in our preliminary analysis of CoMMpass data using genetic ancestry and PCA was a significant difference in the frequency of nonsilent mutations in TP53, with a frequency of 7.1% (33/464) in patients clustering within the European ancestry compared to none (0/142) in African ancestry populations. Further analysis of enrichment of differentially mutated key factors within the TP53 pathway showed ATM as another gene with a significantly (p= 0.019) higher mutation frequency in EA PCA 4.7% (22/464) compared to AA PCA 0.7% (1/142). Analysis of clinical outcomes data showed poorer overall survival in patients harboring TP53 alterations. Furthermore, a comprehensive mutation analysis across samples identified a novel candidate PTCHD3 (p = 7.07E-06) with a significantly higher mutation occurrence in patients of African ancestry. Moreover, the frequency of copy number alterations known to be associated with poor prognosis revealed notable, but not significant (p=0.259) lower frequency of 1q gain in tumors from African compared to European descent. Lastly, we also observed a significant (p=0.0157) two-fold increase in early age of onset of MM in patients of African descent compared to those of European descent. Conclusion: CoMMpass has constructed a fruitful discovery environment at nexus of high-resolution next generation deep sequencing with detailed clinical data allowing to elucidate potential ancestral drivers of MM paving the way to personalized treatments. Ultimately, these data may help us further delineate the influence of percent admixture on biological factors that drive differences in incidence and outcomes among multi-ethnic MM patients. Citation Format: Zarko Manojlovic, Austin Christofferson, Gil Speyer, Seungchan Kim, Winnie Liang, Mary Derome, Daniel Auclair, David Craig, Jonathan Keats, John Carpten. Comprehensive analysis of molecular pathogenesis of multiple myeloma by genetic ancestry [abstract]. In: Proceedings of the AACR International Conference: New Frontiers in Cancer Research; 2017 Jan 18-22; Cape Town, South Africa. Philadelphia (PA): AACR; Cancer Res 2017;77(22 Suppl):Abstract nr A32.

Abstract 1083: Synergistic drug combination prediction through drug differential dependency network analysis

In an effort to discover strategies which identify effective drug combinations, we analyzed 39 of the 480 compounds screened in the Cancer Therapeutics Response Portal (CTRP) where combinations of two compounds were tested against 860 cancer cell lines; this enabled a comparison of the drug sensitivity of the combinations versus that of the individual compounds. More than half of the drug combinations (n=21) did not significantly improve the drug sensitivity, compared to individual compounds alone. In fact, some of the combinations showed reduced drug sensitivity. In EDDY-CTRP* analysis, the Cancer Cell Line Encyclopedia (CCLE) RNAseq data and CTRP compound response measurements were analyzed to discover both 1) pathways enriched with differential dependencies between sensitive and non-sensitive cell lines for each compound and 2) the mediators of cell line response to a drug. A mediator is a gene in a pathway that plays a significantly different role between sensitive and non-sensitive conditions. The significance is assessed for either essentiality, measured as a node’s centrality change, or specificity, measured as the difference in condition specific edges. These drug-pathway-mediator connections are predicted to reveal crucial molecular determinants of drug sensitivity that otherwise are hidden in the complexities of the molecular networks of the cell (Speyer et al., PSB 22:497-508, 2017). We further investigated whether mediators identified for single compounds could predict sensitivity to drug combinations. This analysis revealed that if two single compounds share the same specificity mediators, i.e. the genes with the most significant re-wiring of gene dependencies between sensitive and non-sensitive cell lines, combination of these two compounds correlate with improved sensitivity. The converse was also found: compounds that do not share mediators rarely show synergy. Further analysis and empirical testing of predicted combinations promises to prioritize synergistic drug combinations. We believe that this methodology may predict synergistic drug combinations from cancer cell line drug screening data. Supported by NIH U01CA168397. *available at http://biocomputing.tgen.org/software/EDDY/CTRP Citation Format: Seungchan Kim, Gil Speyer, Harshil Dhruv, Jeff Kiefer, Michael Berens. Synergistic drug combination prediction through drug differential dependency network analysis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 1083. doi:10.1158/1538-7445.AM2017-1083

Abstract 1520: Identifying differential dependency networks accounting for response to NEDD8-inhibitor in large-scale cancer cell line data

The Cancer Cell Line Encyclopedia (CCLE) houses molecular profiles of ∼800 human long term cell lines spanning several different histological types of cancer. In addition, Cancer Therapeutics Response Portal (CTRP) provides drug response measurements for 481 small molecules. Integration of these data enables investigation of the molecular correlates of drug response (sensitivity and resistance). In this current effort, we studied the NEDDylation small molecule inhibitor, MLN4924, in the context of genomic data to uncover novel mechanistic correlates of drug response across the panel of cell lines. We recently reported (Jung and Kim 2016 NAR) development of a robust computational method that shows promise to identify novel insights when applied to multi-dimensional data sets as outlined above. The Evaluation of Differential Dependency (EDDY) employs Bayesian networks to represent statistically distinct differences in relationships between genes within a specific biological pathway as queried between two conditions, in this instance, cell lines that are sensitive and those that are non-sensitive to MLN4924. While EDDY has been successfully employed in the analysis of specific diseases such as TCGA adrenocortical carcinoma, its statistical rigor incurred a prohibitive computational load to assess conditional differences across larger datasets. Recent computational enhancements to EDDY enable processing of larger datasets in reasonable time while maintaining sensitivity. The capability of analyzing broader pan-cancer datasets such as CCLE has enabled EDDY to become more capable in identifying general trends across disease subtypes. Specifically, we demonstrate the enhanced EDDY in analysis of MLN4924 response across the CCLE data set combined with CTRP data set. Initial outcomes from EDDY point to both anticipated and unanticipated biological determinants of response. For example, it is noted that specific oncogenic pathways, such as those centered on PIK3CA, appear to show differential dependencies in the sensitive and non-sensitive cell lines. We also observe genes and candidate pathways related to apoptotic mechanisms that may reveal mechanistic insights to predicting drug response. Specifically, genes and pathways associated with certain apoptotic mechanisms around mitochondrial proteins and glutathione peroxidase may serve as unique determinants of drug response. Multidimensional data analyzed by EDDY uncovers candidate mechanisms of vulnerability to specific small molecule inhibitors, which may guide development of predictive models for treatment planning when using agents with highly context-dependent efficacies. Supported by NIH U01CA168397 Citation Format: Gil Speyer, Harshil Dhruv, Jeff Kiefer, Stuart Schreiber, Paul Clemons, Michael E. Berens, Seungchan Kim. Identifying differential dependency networks accounting for response to NEDD8-inhibitor in large-scale cancer cell line data. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 1520.