Citations

Background Identifying the biologic functions of groups of genes identified in high-throughput studies currently requires considerable time and/or bioinformatics experience. This is due in part to each resource housed within separate databases, requiring users to know about them, and integrate across them. Time consuming and often repeated for each study, integrating across resources and merging with data under study is an increasingly common bioinformatics task. Methods We developed an open-source R software package for assisting researchers in annotating their genesets with functions, pathways, and their interconnectivity across a diversity of network resources. Results We present rapid integration of term annotation and network resources (RITAN) for the rapid and comprehensive annotation of a list of genes using functional term and pathway resources and their relationships among each other using multiple network biology resources. Currently, and to comply with data redistribution policies, RITAN allows rapid access to 16 term annotations spanning gene ontology, biologic pathways, and immunologic modules, and nine network biology resources, with support for user-supplied resources; we provide recommendations for additional resources and scripts to facilitate their addition to RITAN. Having the resources together in the same system allows users to derive novel combinations. RITAN has a growing set of tools to explore the relationships within resources themselves. These tools allow users to merge resources together such that the merged annotations have a minimal overlap with one another. Because we index both function annotation and network interactions, the combination allows users to expand small groups of genes using links from biologic networks—either by adding all neighboring genes or by identifying genes that efficiently connect among input genes—followed by term enrichment to identify functions. That is, users can start from a core set of genes, identify interacting genes from biologic networks, and then identify the functions to which the expanded list of genes contribute. Conclusion We believe RITAN fills the important niche of bridging the results of high-throughput experiments with the ever-growing corpus of functional annotations and network biology resources. Availability Rapid integration of term annotation and network resources is available as an R package at github.com/MTZimmer/RITAN and BioConductor.org .

Background In the phase 3 MARIANNE trial, trastuzumab emtansine (T‐DM1) with or without pertuzumab showed noninferior progression‐free survival and better tolerability than trastuzumab plus a taxane (HT) for the first‐line treatment of human epidermal growth factor receptor 2 (HER2)–positive advanced breast cancer. This article reports the final descriptive overall survival (OS) analysis, updated safety data, and additional patient‐reported outcomes and biomarker analyses. Methods OS was assessed in 1095 patients with HER2‐positive breast cancer and no prior therapy for advanced disease who had been randomized to HT, T‐DM1 plus a placebo (hereafter T‐DM1), or T‐DM1 plus pertuzumab (T‐DM1+pertuzumab). A post hoc exploratory landmark analysis of OS, baseline patient and disease characteristics, and tumor biomarkers in patients with and without an objective tumor response (OR) according to the Response Evaluation Criteria in Solid Tumors within 6.5 months of randomization was conducted. Results The median OS was similar across groups (50.9, 53.7, and 51.8 months for the HT, T‐DM1, and T‐DM1+pertuzumab groups, respectively). Among patients with an OR, the median OS was longer with T‐DM1 (64.4 months) and T‐DM1+pertuzumab (not reached) versus HT (56.3 months). No baseline characteristics or biomarkers were strongly associated with OR. The incidence of grade 3 or higher adverse events was greater with HT (55.8%) than T‐DM1 (47.1%) or T‐DM1+pertuzumab (48.6%). The median time to clinically meaningful deterioration (a 3‐point or greater change) in neurotoxicity symptoms was shorter with HT (2.1 months) and T‐DM1+pertuzumab (4.2 months) than T‐DM1 (6.2 months). Fewer patients reported alopecia and diarrhea and were bothered by treatment side effects in the T‐DM1 arm. Conclusions These results support T‐DM1 as a first‐line treatment for patients with HER2‐positive metastatic breast cancer who are deemed unsuitable for taxane‐based therapy.

Abstract Purpose: Ovarian cancer is a heterogeneous disease that can be divided into multiple subtypes with variable etiology, pathogenesis, and prognosis. We analyzed DNA methylation profiling data to identify biologic subgroups of ovarian cancer and study their relationship with histologic subtypes, copy number variation, RNA expression data, and outcomes. Experimental Design: A total of 162 paraffin-embedded ovarian epithelial tumor tissues, including the five major epithelial ovarian tumor subtypes (high- and low-grade serous, endometrioid, mucinous, and clear cell) and tumors of low malignant potential were selected from two different sources: The Polish Ovarian Cancer study, and the Surveillance, Epidemiology, and End Results Residual Tissue Repository (SEER RTR). Analyses were restricted to Caucasian women. Methylation profiling was conducted using the Illumina 450K methylation array. For 45 tumors array copy number data were available. NanoString gene expression data for 39 genes were available for 61 high-grade serous carcinomas (HGSC). Results: Consensus nonnegative matrix factorization clustering of the 1,000 most variable CpG sites showed four major clusters among all epithelial ovarian cancers. We observed statistically significant differences in survival (log-rank test, P = 9.1 × 10−7) and genomic instability across these clusters. Within HGSC, clustering showed three subgroups with survival differences (log-rank test, P = 0.002). Comparing models with and without methylation subgroups in addition to previously identified gene expression subtypes suggested that the methylation subgroups added significant survival information (P = 0.007). Conclusions: DNA methylation profiling of ovarian cancer identified novel molecular subgroups that had significant survival difference and provided insights into the molecular underpinnings of ovarian cancer. See related commentary by Ishak et al., p. 5729

Abstract Humans vary considerably both in their baseline and activated immune phenotypes. We developed a user-friendly open-access web portal, ImmuneRegulation, that enables users to interactively explore immune regulatory elements that drive cell-type or cohort-specific gene expression levels. ImmuneRegulation currently provides the largest centrally integrated resource on human transcriptome regulation across whole blood and blood cell types, including (i) ∼43,000 genotyped individuals with associated gene expression data from ∼51,000 experiments, yielding genetic variant-gene expression associations on ∼220 million eQTLs; (ii) 14 million transcription factor (TF)-binding region hits extracted from 1945 ChIP-seq studies; and (iii) the latest GWAS catalog with 67,230 published variant-trait associations. Users can interactively explore associations between queried gene(s) and their regulators (cis-eQTLs, trans-eQTLs or TFs) across multiple cohorts and studies. These regulators may explain genotype-dependent gene expression variations and be critical in selecting the ideal cohorts or cell types for follow-up studies or in developing predictive models. Overall, ImmuneRegulation significantly lowers the barriers between complex immune regulation data and researchers who want rapid, intuitive and high-quality access to the effects of regulatory elements on gene expression in multiple studies to empower investigators in translating these rich data into biological insights and clinical applications, and is freely available at https://immuneregulation.mssm.edu.

Vaccination is an effective prevention of influenza infection. However, certain individuals develop a lower antibody response after vaccination, which may lead to susceptibility to subsequent infection. An important challenge in human health is to find baseline gene signatures to help identify individuals who are at higher risk for infection despite influenza vaccination. We developed a multi-level machine learning strategy to build a predictive model of vaccine response using pre−vaccination antibody titers and network interactions between pre−vaccination gene expression levels. The first-level baseline−antibody model explains a significant amount of variation in post-vaccination response, especially for subjects with large pre−existing antibody titers. In the second level, we clustered individuals based on pre−vaccination antibody titers to focus gene−based modeling on individuals with lower baseline HAI where additional response variation may be predicted by baseline gene expression levels. In the third level, we used a gene−association interaction network (GAIN) feature selection algorithm to find the best pairs of genes that interact to influence antibody response within each baseline titer cluster. We used ratios of the top interacting genes as predictors to stabilize machine learning model generalizability. We trained and tested the multi-level approach on data with young and older individuals immunized against influenza vaccine in multiple cohorts. Our results indicate that the GAIN feature selection approach improves model generalizability and identifies genes enriched for immunologically relevant pathways, including B Cell Receptor signaling and antigen processing. Using a multi-level approach, starting with a baseline HAI model and stratifying on baseline HAI, allows for more targeted gene−based modeling. We provide an interactive tool that may be extended to other vaccine studies.