Extracellular Vesicle - MDM2-DNA as a Potential Liquid Biopsy Biomarker for Disease Identification in Retroperitoneal Liposarcoma

OBJECTIVE

We aimed to assess the levels of MDM2-DNA within extracellular vesicles (EVs) isolated from the serum of retroperitoneal liposarcoma (RLS) patients versus healthy donors, as well as within the same patients at the time of surgery versus post-operative surveillance visits. To determine whether EV-MDM2 may serve as a possible first-ever biomarker of liposarcoma recurrence.

BACKGROUND

A hallmark of well-differentiated and de-differentiated (WD/DD) retroperitoneal liposarcoma is elevated MDM2 due to genome amplification, with recurrence rates of >50% even after complete resection. Imaging technologies frequently cannot resolve recurrent WD/DD-RLS versus postoperative scarring. Early detection of recurrent lesions, for which biomarkers are lacking, would guide surveillance and treatment decisions.

METHODS

WD/DD-RLS serum samples were collected both at the time of surgery and during follow-up visits from 42 patients, along with sera from healthy donors (n=14). EVs were isolated, DNA purified and MDM2-DNA levels determined through q-PCR analysis. Non-parametric tests were employed to compare EV-MDM2 DNA levels from patients versus control group, as well as the time of surgery versus post-surgery conditions.

RESULTS

EV-MDM2 levels were significantly higher in WD/DD-RLS than controls (P= 0.00085). Moreover, EV-MDM2 levels were remarkably decreased in WD/DD-RLS patients after resection (P=0.00036), reaching values comparable to control group (P=0.124). During post-operative surveillance, significant increases of EV-MDM2 was observed in some patients, correlating with CT scan evidence of recurrent or persistent post-resection disease.

CONCLUSIONS

Serum EV-MDM2 may serve as a potential biomarker of early recurrent or post-operatively persistent WD/DD-RLS, a disease currently lacking such determinants.

Transcriptome Complexity Disentangled: A Regulatory Molecules Approach

Transcription factors (TFs) and microRNAs (miR-NAs) are fundamental regulators of gene expression, cell state, and biological processes. This study investigated whether a small subset of TFs and miRNAs could accurately predict genome-wide gene expression. We analyzed 8895 samples across 31 cancer types from The Cancer Genome Atlas and identified 28 miRNA and 28 TF clusters using unsupervised learning. Medoids of these clusters could differentiate tissues of origin with 92.8% accuracy, demonstrating their biological relevance. We developed Tissue-Agnostic and Tissue-Aware models to predict 20,000 gene expressions using the 56 selected medoid miR-NAs and TFs. The Tissue-Aware model attained an R 2 of 0.70 by incorporating tissue-specific information. Despite measuring only 1/400th of the transcriptome, the prediction accuracy was comparable to that achieved by the 1000 landmark genes. This suggests the transcriptome has an intrinsically low-dimensional structure that can be captured by a few regulatory molecules. Our approach could enable cheaper transcriptome assays and analysis of low-quality samples. It also provides insights into genes that are heavily regulated by miRNAs/TFs versus alternative mechanisms. However, model transportability was impacted by dataset discrepancies, especially in miRNA distribution. Overall, this study demonstrates the potential of a biology-guided approach for robust transcriptome representation.

SillyPutty: Improved clustering by optimizing the silhouette width

Unsupervised clustering is an important task in biomedical science. We developed a new clustering method, called SillyPutty, for unsupervised clustering. As test data, we generated a series of datasets using the Umpire R package. Using these datasets, we compared SillyPutty to several existing algorithms using multiple metrics (Silhouette Width, Adjusted Rand Index, Entropy, Normalized Within-group Sum of Square errors, and Perfect Classification Count). Our findings revealed that SillyPutty is a valid standalone clustering method, comparable in accuracy to the best existing methods. We also found that the combination of hierarchical clustering followed by SillyPutty has the best overall performance in terms of both accuracy and speed.

Oncogenic role and target properties of the lysine-specific demethylase KDM1A in chronic lymphocytic leukemia

In chronic lymphocytic leukemia (CLL), epigenetic alterations are considered to centrally shape the transcriptional signatures that drive disease evolution and underlie its biological and clinical subsets. Characterizations of epigenetic regulators, particularly histone-modifying enzymes, are very rudimentary in CLL. In efforts to establish effectors of the CLL-associated oncogene T-cell leukemia 1A (TCL1A), we identified here the lysine-specific histone demethylase KDM1A to interact with the TCL1A protein in B cells in conjunction with an increased catalytic activity of KDM1A. We demonstrate that KDM1A is upregulated in malignant B cells. Elevated KDM1A and associated gene expression signatures correlated with aggressive disease features and adverse clinical outcomes in a large prospective CLL trial cohort. Genetic Kdm1a knockdown in Eμ-TCL1A mice reduced leukemic burden and prolonged animal survival, accompanied by upregulated p53 and proapoptotic pathways. Genetic KDM1A depletion also affected milieu components (T, stromal, and monocytic cells), resulting in significant reductions in their capacity to support CLL-cell survival and proliferation. Integrated analyses of differential global transcriptomes (RNA sequencing) and H3K4me3 marks (chromatin immunoprecipitation sequencing) in Eμ-TCL1A vs iKdm1aKD;Eμ-TCL1A mice (confirmed in human CLL) implicate KDM1A as an oncogenic transcriptional repressor in CLL which alters histone methylation patterns with pronounced effects on defined cell death and motility pathways. Finally, pharmacologic KDM1A inhibition altered H3K4/9 target methylation and revealed marked anti-B-cell leukemic synergisms. Overall, we established the pathogenic role and effector networks of KDM1A in CLL via tumor-cell intrinsic mechanisms and its impacts in cells of the microenvironment. Our data also provide rationales to further investigate therapeutic KDM1A targeting in CLL.

Intratumoral heterogeneity and clonal evolution induced by HPV integration

The human papillomavirus (HPV) genome is integrated into host DNA in most HPV-positive cancers, but the consequences for chromosomal integrity are unknown. Continuous long-read sequencing of oropharyngeal cancers and cancer cell lines identified a previously undescribed form of structural variation, "heterocateny," characterized by diverse, interrelated, and repetitive patterns of concatemerized virus and host DNA segments within a cancer. Unique breakpoints shared across structural variants facilitated stepwise reconstruction of their evolution from a common molecular ancestor. This analysis revealed that virus and virus-host concatemers are unstable and, upon insertion into and excision from chromosomes, facilitate capture, amplification, and recombination of host DNA and chromosomal rearrangements. Evidence of heterocateny was detected in extrachromosomal and intrachromosomal DNA. These findings indicate that heterocateny is driven by the dynamic, aberrant replication and recombination of an oncogenic DNA virus, thereby extending known consequences of HPV integration to include promotion of intratumoral heterogeneity and clonal evolution.

CEDA: integrating gene expression data with CRISPR-pooled screen data identifies essential genes with higher expression

MOTIVATION

Clustered regularly interspaced short palindromic repeats (CRISPR)-based genetic perturbation screen is a powerful tool to probe gene function. However, experimental noises, especially for the lowly expressed genes, need to be accounted for to maintain proper control of false positive rate.

METHODS

We develop a statistical method, named CRISPR screen with Expression Data Analysis (CEDA), to integrate gene expression profiles and CRISPR screen data for identifying essential genes. CEDA stratifies genes based on expression level and adopts a three-component mixture model for the log-fold change of single-guide RNAs (sgRNAs). Empirical Bayesian prior and expectation-maximization algorithm are used for parameter estimation and false discovery rate inference.

RESULTS

Taking advantage of gene expression data, CEDA identifies essential genes with higher expression. Compared to existing methods, CEDA shows comparable reliability but higher sensitivity in detecting essential genes with moderate sgRNA fold change. Therefore, using the same CRISPR data, CEDA generates an additional hit gene list.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Sequences of Events from the Electronic Medical Record and the Onset of Infection

We present a novel model of time-series analysis to learn from electronic health record (EHR) data when infection occurred in the intensive care unit (ICU) by translating methods from proteomics and Bayesian statistics. Using 48,536 patients hospitalized in an ICU, we describe each hospital course as an 'alphabet' of 23 physician actions ('events') in temporal order. We analyze these as k-mers of length 3-12 events and apply a Bayesian model of (cumulative) relative risk (RR). The log2-transformed RR (median=0.248, mean=0.226) supported the conclusion that the events selected were individually associated with increased risk of infection. Selecting from all possible cutoffs of maximum gain (MG), MG>0.0244 predicts administration of antibiotics with PPV 82.0 %, NPV 44.4 %, and AUC 0.706. Our approach holds value for retrospective analysis of other clinical syndromes for which time-of-onset is critical to analysis but poorly marked in EHRs, including delirium and decompensation.

Diverse tumorigenic consequences of human papillomavirus integration in primary oropharyngeal cancers

Human papillomavirus (HPV) causes 5% of all cancers and frequently integrates into host chromosomes. The HPV oncoproteins E6 and E7 are necessary but insufficient for cancer formation, indicating that additional secondary genetic events are required. Here, we investigate potential oncogenic impacts of virus integration. Analysis of 105 HPV-positive oropharyngeal cancers by whole-genome sequencing detects virus integration in 77%, revealing five statistically significant sites of recurrent integration near genes that regulate epithelial stem cell maintenance (i.e., SOX2, TP63, FGFR, MYC) and immune evasion (i.e., CD274). Genomic copy number hyperamplification is enriched 16-fold near HPV integrants, and the extent of focal host genomic instability increases with their local density. The frequency of genes expressed at extreme outlier levels is increased 86-fold within ±150 kb of integrants. Across 95% of tumors with integration, host gene transcription is disrupted via intragenic integrants, chimeric transcription, outlier expression, gene breaking, and/or de novo expression of noncoding or imprinted genes. We conclude that virus integration can contribute to carcinogenesis in a large majority of HPV-positive oropharyngeal cancers by inducing extensive disruption of host genome structure and gene expression.

RCytoGPS: An R Package for Reading and Visualizing Cytogenetics Data

SUMMARY

Cytogenetics data, or karyotypes, are among the most common clinically used forms of genetic data. Karyotypes are stored as standardized text strings using the International System for Human Cytogenomic Nomenclature (ISCN). Historically, these data have not been used in large-scale computational analyses due to limitations in the ISCN text format and structure. Recently developed computational tools such as CytoGPS have enabled large-scale computational analyses of karyotypes. To further enable such analyses, we have now developed RCytoGPS, an R package that takes JSON files generated from CytoGPS.org and converts them into objects in R. This conversion facilitates the analysis and visualizations of karyotype data. In effect this tool streamlines the process of performing large-scale karyotype analyses, thus advancing the field of computational cytogenetic pathology.

AVAILABILITY AND IMPLEMENTATION

Freely available at https://CRAN.R-project.org/package=RCytoGPS. The code for the underlying CytoGPS software can be found at https://github.com/i2-wustl/CytoGPS.

SUPPLEMENTARY INFORMATION

There is no supplementary data.

Simulation-derived best practices for clustering clinical data

INTRODUCTION

Clustering analyses in clinical contexts hold promise to improve the understanding of patient phenotype and disease course in chronic and acute clinical medicine. However, work remains to ensure that solutions are rigorous, valid, and reproducible. In this paper, we evaluate best practices for dissimilarity matrix calculation and clustering on mixed-type, clinical data.

METHODS

We simulate clinical data to represent problems in clinical trials, cohort studies, and EHR data, including single-type datasets (binary, continuous, categorical) and 4 data mixtures. We test 5 single distance metrics (Jaccard, Hamming, Gower, Manhattan, Euclidean) and 3 mixed distance metrics (DAISY, Supersom, and Mercator) with 3 clustering algorithms (hierarchical (HC), k-medoids, self-organizing maps (SOM)). We quantitatively and visually validate by Adjusted Rand Index (ARI) and silhouette width (SW). We applied our best methods to two real-world data sets: (1) 21 features collected on 247 patients with chronic lymphocytic leukemia, and (2) 40 features collected on 6000 patients admitted to an intensive care unit.

RESULTS

HC outperformed k-medoids and SOM by ARI across data types. DAISY produced the highest mean ARI for mixed data types for all mixtures except unbalanced mixtures dominated by continuous data. Compared to other methods, DAISY with HC uncovered superior, separable clusters in both real-world data sets.

DISCUSSION

Selecting an appropriate mixed-type metric allows the investigator to obtain optimal separation of patient clusters and get maximum use of their data. Superior metrics for mixed-type data handle multiple data types using multiple, type-focused distances. Better subclassification of disease opens avenues for targeted treatments, precision medicine, clinical decision support, and improved patient outcomes.

DNA methylation epitypes highlight underlying developmental and disease pathways in acute myeloid leukemia

Acute myeloid leukemia (AML) is a molecularly complex disease characterized by heterogeneous tumor genetic profiles and involving numerous pathogenic mechanisms and pathways. Integration of molecular data types across multiple patient cohorts may advance current genetic approaches for improved subclassification and understanding of the biology of the disease. Here, we analyzed genome-wide DNA methylation in 649 AML patients using Illumina arrays and identified a configuration of 13 subtypes (termed “epitypes”) using unbiased clustering. Integration of genetic data revealed that most epitypes were associated with a certain recurrent mutation (or combination) in a majority of patients, yet other epitypes were largely independent. Epitypes showed developmental blockage at discrete stages of myeloid differentiation, revealing epitypes that retain arrested hematopoietic stem-cell-like phenotypes. Detailed analyses of DNA methylation patterns identified unique patterns of aberrant hyper- and hypomethylation among epitypes, with variable involvement of transcription factors influencing promoter, enhancer, and repressed regions. Patients in epitypes with stem-cell-like methylation features showed inferior overall survival along with up-regulated stem cell gene expression signatures. We further identified a DNA methylation signature involving STAT motifs associated with FLT3-ITD mutations. Finally, DNA methylation signatures were stable at relapse for the large majority of patients, and rare epitype switching accompanied loss of the dominant epitype mutations and reversion to stem-cell-like methylation patterns. These results show that DNA methylation-based classification integrates important molecular features of AML to reveal the diverse pathogenic and biological aspects of the disease.

A novel model to label delirium in an intensive care unit from clinician actions

BACKGROUND

In the intensive care unit (ICU), delirium is a common, acute, confusional state associated with high risk for short- and long-term morbidity and mortality. Machine learning (ML) has promise to address research priorities and improve delirium outcomes. However, due to clinical and billing conventions, delirium is often inconsistently or incompletely labeled in electronic health record (EHR) datasets. Here, we identify clinical actions abstracted from clinical guidelines in electronic health records (EHR) data that indicate risk of delirium among intensive care unit (ICU) patients. We develop a novel prediction model to label patients with delirium based on a large data set and assess model performance.

METHODS

EHR data on 48,451 admissions from 2001 to 2012, available through Medical Information Mart for Intensive Care-III database (MIMIC-III), was used to identify features to develop our prediction models. Five binary ML classification models (Logistic Regression; Classification and Regression Trees; Random Forests; Naïve Bayes; and Support Vector Machines) were fit and ranked by Area Under the Curve (AUC) scores. We compared our best model with two models previously proposed in the literature for goodness of fit, precision, and through biological validation.

RESULTS

Our best performing model with threshold reclassification for predicting delirium was based on a multiple logistic regression using the 31 clinical actions (AUC 0.83). Our model out performed other proposed models by biological validation on clinically meaningful, delirium-associated outcomes.

CONCLUSIONS

Hurdles in identifying accurate labels in large-scale datasets limit clinical applications of ML in delirium. We developed a novel labeling model for delirium in the ICU using a large, public data set. By using guideline-directed clinical actions independent from risk factors, treatments, and outcomes as model predictors, our classifier could be used as a delirium label for future clinically targeted models.

Umpire 2.0: Simulating realistic, mixed-type, clinical data for machine learning

The Umpire 2.0 R-package offers a streamlined, user-friendly workflow to simulate complex, heterogeneous, mixed-type data with known subgroup identities, dichotomous outcomes, and time-to-event data, while providing ample opportunities for fine-tuning and flexibility. Here, we describe how we have expanded the core Umpire 1.0 R-package, developed to simulate gene expression data, to generate clinically realistic, mixed-type data for use in evaluating unsupervised and supervised machine learning (ML) methods. As the availability of large-scale clinical data for ML has increased, clinical data has posed unique challenges, including widely variable size, individual biological heterogeneity, data collection and measurement noise, and mixed data types. Developing and validating ML methods for clinical data requires data sets with known ground truth, generated from simulation. Umpire 2.0 addresses challenges to simulating realistic clinical data by providing the user a series of modules to generate survival parameters and subgroups, apply meaningful additive noise, and discretize to single or mixed data types. Umpire 2.0 provides broad functionality across sample sizes, feature spaces, and data types, allowing the user to simulate correlated, heterogeneous, binary, continuous, categorical, or mixed type data from the scale of a small clinical trial to data on thousands of patients drawn from electronic health records. The user may generate elaborate simulations by varying parameters in order to compare algorithms or interrogate operating characteristics of an algorithm in both supervised and unsupervised ML.

Pattern recognition in lymphoid malignancies using CytoGPS and Mercator

BACKGROUND

There have been many recent breakthroughs in processing and analyzing large-scale data sets in biomedical informatics. For example, the CytoGPS algorithm has enabled the use of text-based karyotypes by transforming them into a binary model. However, such advances are accompanied by new problems of data sparsity, heterogeneity, and noisiness that are magnified by the large-scale multidimensional nature of the data. To address these problems, we developed the Mercator R package, which processes and visualizes binary biomedical data. We use Mercator to address biomedical questions of cytogenetic patterns relating to lymphoid hematologic malignancies, which include a broad set of leukemias and lymphomas. Karyotype data are one of the most common form of genetic data collected on lymphoid malignancies, because karyotyping is part of the standard of care in these cancers.

RESULTS

In this paper we combine the analytic power of CytoGPS and Mercator to perform a large-scale multidimensional pattern recognition study on 22,741 karyotype samples in 47 different hematologic malignancies obtained from the public Mitelman database.

CONCLUSION

Our findings indicate that Mercator was able to identify both known and novel cytogenetic patterns across different lymphoid malignancies, furthering our understanding of the genetics of these diseases.

Mercator: A Pipeline For Multi-Method, Unsupervised Visualization And Distance Generation

Summary

Unsupervised machine learning provides tools for researchers to uncover latent patterns in large-scale data, based on calculated distances between observations. Methods to visualize high-dimensional data based on these distances can elucidate subtypes and interactions within multi-dimensional and high-throughput data. However, researchers can select from a vast number of distance metrics and visualizations, each with their own strengths and weaknesses. The Mercator R package facilitates selection of a biologically meaningful distance from 10 metrics, together appropriate for binary, categorical and continuous data, and visualization with 5 standard and high-dimensional graphics tools. Mercator provides a user-friendly pipeline for informaticians or biologists to perform unsupervised analyses, from exploratory pattern recognition to production of publication-quality graphics.

Availabilityand implementation

Mercator is freely available at the Comprehensive R Archive Network (https://cran.r-project.org/web/packages/Mercator/index.html).

CytoGPS: A large-scale karyotype analysis of CML data

Karyotyping, the practice of visually examining and recording chromosomal abnormalities, is commonly used to diagnose diseases of genetic origin, including cancers. Karyotypes are recorded as text written in the International System for Human Cytogenetic Nomenclature (ISCN). Downstream analysis of karyotypes is conducted manually, due to the visual nature of analysis and the linguistic structure of the ISCN. The ISCN has not been computer-readable and, as such, prevents the full potential of these genomic data from being realized. In response, we developed CytoGPS, a platform to analyze large volumes of cytogenetic data using a Loss-Gain-Fusion model that converts the human-readable ISCN karyotypes into a machine-readable binary format. As proof of principle, we applied CytoGPS to cytogenetic data from the Mitelman Database of Chromosome Aberrations and Gene Fusions in Cancer, a National Cancer Institute hosted database of over 69,000 karyotypes of human cancers. Using the Jaccard coefficient to determine similarity between karyotypes structured as binary vectors, we were able to identify novel patterns from 4,968 Mitelman CML karyotypes, such as the co-occurrence of trisomy 19 and 21. The CytoGPS platform unlocks the potential for large-scale, comparative analysis of cytogenetic data. This methodological platform is freely available at CytoGPS.org.

Unsupervised machine learning and prognostic factors of survival in chronic lymphocytic leukemia

OBJECTIVE

Unsupervised machine learning approaches hold promise for large-scale clinical data. However, the heterogeneity of clinical data raises new methodological challenges in feature selection, choosing a distance metric that captures biological meaning, and visualization. We hypothesized that clustering could discover prognostic groups from patients with chronic lymphocytic leukemia, a disease that provides biological validation through well-understood outcomes.

METHODS

To address this challenge, we applied k-medoids clustering with 10 distance metrics to 2 experiments ("A" and "B") with mixed clinical features collapsed to binary vectors and visualized with both multidimensional scaling and t-stochastic neighbor embedding. To assess prognostic utility, we performed survival analysis using a Cox proportional hazard model, log-rank test, and Kaplan-Meier curves.

RESULTS

In both experiments, survival analysis revealed a statistically significant association between clusters and survival outcomes (A: overall survival, P = .0164; B: time from diagnosis to treatment, P = .0039). Multidimensional scaling separated clusters along a gradient mirroring the order of overall survival. Longer survival was associated with mutated immunoglobulin heavy-chain variable region gene (IGHV) status, absent Zap 70 expression, female sex, and younger age.

CONCLUSIONS

This approach to mixed-type data handling and selection of distance metric captured well-understood, binary, prognostic markers in chronic lymphocytic leukemia (sex, IGHV mutation status, ZAP70 expression status) with high fidelity.

CytoGPS: a web-enabled karyotype analysis tool for cytogenetics

Summary

Karyotype data are the most common form of genetic data that is regularly used clinically. They are collected as part of the standard of care in many diseases, particularly in pediatric and cancer medicine contexts. Karyotypes are represented in a unique text-based format, with a syntax defined by the International System for human Cytogenetic Nomenclature (ISCN). While human-readable, ISCN is not intrinsically machine-readable. This limitation has prevented the full use of complex karyotype data in discovery science use cases. To enhance the utility and value of karyotype data, we developed a tool named CytoGPS. CytoGPS first parses ISCN karyotypes into a machine-readable format. It then converts the ISCN karyotype into a binary Loss-Gain-Fusion (LGF) model, which represents all cytogenetic abnormalities as combinations of loss, gain, or fusion events, in a format that is analyzable using modern computational methods. Such data is then made available for comprehensive ‘downstream’ analyses that previously were not feasible.

Availability and implementation

Freely available at http://cytogps.org.

LC-FACSeq is a method for detecting rare clones in leukemia

Detecting, characterizing, and monitoring rare populations of cells can increase testing sensitivity, give insight into disease mechanism, and inform clinical decision making. One area that can benefit from increased resolution is management of cancers in clinical remission but with measurable residual disease (MRD) by multicolor FACS. Detecting and monitoring genomic clonal resistance to treatment in the setting of MRD is technically difficult and resource intensive due to the limited amounts of disease cells. Here, we describe limited-cell FACS sequencing (LC-FACSeq), a reproducible, highly sensitive method of characterizing clonal evolution in rare cells relevant to different types of acute and chronic leukemias. We demonstrate the utility of LC-FACSeq for broad multigene gene panels and its application for monitoring sequential acquisition of mutations conferring therapy resistance and clonal evolution in long-term ibrutinib treatment of patients with chronic lymphocytic leukemia. This technique is generalizable for monitoring of other blood and marrow infiltrating cancers.

Explaining Gene Expression Using Twenty-One MicroRNAs

The transcriptome of a tumor contains detailed information about the disease. Although advances in sequencing technologies have generated larger data sets, there are still many questions about exactly how the transcriptome is regulated. One class of regulatory elements consists of microRNAs (or miRs), many of which are known to be associated with cancer. To better understand the relationships between miRs and cancers, we analyzed ∼9000 samples from 32 cancer types studied in The Cancer Genome Atlas. Our feature reduction algorithm found evidence for 21 biologically interpretable clusters of miRs, many of which were statistically associated with a specific type of cancer. Moreover, the clusters contain sufficient information to distinguish between most types of cancer. We then used linear models to measure, genome-wide, how much variation in gene expression could be explained by the 21 average expression values ("scores") of the clusters. Based on the ∼20,000 per-gene R² values, we found that (1) mean differences between tissues of origin explain about 36% of variation; (2) the 21 miR cluster scores explain about 30% of the variation; and (3) combining tissue type with the miR scores explained about 56% of the total genome-wide variation in gene expression. Our analysis of poorly explained genes shows that they are enriched for olfactory receptor processes, sensory perception, and nervous system processing, which are necessary to receive and interpret signals from outside the organism. Therefore, it is reasonable for those genes to be always active and not get downregulated by miRs. In contrast, highly explained genes are characterized by genes translating to proteins necessary for transport, plasma membrane, or metabolic processes that are heavily regulated processes inside the cell. Other genetic regulatory elements such as transcription factors and methylation might help explain some of the remaining variation in gene expression.

Time-to-progression after front-line fludarabine, cyclophosphamide, and rituximab chemoimmunotherapy for chronic lymphocytic leukaemia: a retrospective, multicohort study

BACKGROUND

Fludarabine, cyclophosphamide, and rituximab (FCR) has become a gold-standard chemoimmunotherapy regimen for patients with chronic lymphocytic leukaemia. However, the question remains of how to treat treatment-naive patients with IGHV-unmutated chronic lymphocytic leukaemia. We therefore aimed to develop and validate a gene expression signature to identify which of these patients are likely to achieve durable remissions with FCR chemoimmunotherapy.

METHODS

We did a retrospective cohort study in two cohorts of treatment-naive patients (aged ≥18 years) with chronic lymphocytic leukaemia. The discovery and training cohort consisted of peripheral blood samples collected from patients treated at the University of Texas MD Anderson Cancer Center (Houston, TX, USA), who fulfilled the diagnostic criteria of the International Workshop on Chronic Lymphocytic Leukemia, had received at least three cycles of FCR chemoimmunotherapy, and had been treated between Oct 10, 2000, and Oct 26, 2006 (ie, the MDACC cohort). We did transcriptional profiling on samples obtained from the MDACC cohort to identify genes associated with time to progression. We did univariate Cox proportional hazards analyses and used significant genes to cluster IGHV-unmutated samples into two groups (intermediate prognosis and unfavourable prognosis). After using cross-validation to assess robustness, we applied the Lasso method to standardise the gene expression values to find a minimum gene signature. We validated this signature in an external cohort of treatment-naive patients with IGHV-unmutated chronic lymphocytic leukaemia enrolled on the CLL8 trial of the German Chronic Lymphocytic Leukaemia Study Group who were treated between July 21, 2003, and April 4, 2006 (ie, the CLL8 cohort).

FINDINGS

The MDACC cohort consisted of 101 patients and the CLL8 cohort consisted of 109 patients. Using the MDACC cohort, we identified and developed a 17-gene expression signature that distinguished IGHV-unmutated patients who were likely to achieve a long-term remission following front-line FCR chemoimmunotherapy from those who might benefit from alternative front-line regimens (hazard ratio 3·83, 95% CI 1·94-7·59; p<0·0001). We validated this gene signature in the CLL8 cohort; patients with an unfavourable prognosis versus those with an intermediate prognosis had a cause-specific hazard ratio of 1·90 (95% CI 1·18-3·06; p=0·008). Median time to progression was 39 months (IQR 22-69) for those with an unfavourable prognosis compared with 59 months (28-84) for those with an intermediate prognosis.

INTERPRETATION

We have developed a robust, reproducible 17-gene signature that identifies a subset of treatment-naive patients with IGHV-unmutated chronic lymphocytic leukaemia who might substantially benefit from treatment with FCR chemoimmunotherapy. We recommend testing the value of this gene signature in a prospective study that compares FCR treatment with newer alternative therapies as part of a randomised clinical trial.

FUNDING

Chronic Lymphocytic Leukaemia Global Research Foundation and the National Institutes of Health/National Cancer Institute.

Inferring clonal heterogeneity in cancer using SNP arrays and whole genome sequencing

MOTIVATION

Clonal heterogeneity is common in many types of cancer, including chronic lymphocytic leukemia (CLL). Previous research suggests that the presence of multiple distinct cancer clones is associated with clinical outcome. Detection of clonal heterogeneity from high throughput data, such as sequencing or single nucleotide polymorphism (SNP) array data, is important for gaining a better understanding of cancer and may improve prediction of clinical outcome or response to treatment. Here, we present a new method, CloneSeeker, for inferring clinical heterogeneity from sequencing data, SNP array data, or both.

RESULTS

We generated simulated SNP array and sequencing data and applied CloneSeeker along with two other methods. We demonstrate that CloneSeeker is more accurate than existing algorithms at determining the number of clones, distribution of cancer cells among clones, and mutation and/or copy numbers belonging to each clone. Next, we applied CloneSeeker to SNP array data from samples of 258 previously untreated CLL patients to gain a better understanding of the characteristics of CLL tumors and to elucidate the relationship between clonal heterogeneity and clinical outcome. We found that a significant majority of CLL patients appear to have multiple clones distinguished by copy number alterations alone. We also found that the presence of multiple clones corresponded with significantly worse survival among CLL patients. These findings may prove useful for improving the accuracy of prognosis and design of treatment strategies.

AVAILABILITY AND IMPLEMENTATION

Code available on R-Forge: https://r-forge.r-project.org/projects/CloneSeeker/.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

LGALS3 is connected to CD74 in a previously unknown protein network that is associated with poor survival in patients with AML

Background

Galectin 3 (LGALS3) gene expression is associated with poor survival in acute myeloid leukemia (AML) but the prognostic impact of LGALS3 protein expression in AML is unknown. LGALS3 supports diverse survival pathways including RAS mediated cascades, protein expression and stability of anti-apoptotic BCL2 family members, and activation of proliferative pathways including those mediated by beta Catenin. CD74 is a positive regulator of CD44 and CXCR4 signaling and this molecule may be critical for AML stem cell function. At present, the role of LGALS3 and CD74 in AML is unclear. In this study, we examine protein expression of LGALS3 and CD74 by reverse phase protein analysis (RPPA) and identify new protein networks associated with these molecules. In addition, we determine prognostic potential of LGALS3, CD74, and their protein networks for clinical correlates in AML patients.

Methods

RPPA was used to determine relative expression of LGALS3, CD74, and 229 other proteins in 231 fresh AML patient samples and 205 samples were from patients who were treated and evaluable for outcome. Pearson correlation analysis was performed to identify proteins associated with LGALS3 and CD74. Progeny clustering was performed to generate protein networks. String analysis was performed to determine protein:protein interactions in networks and to perform gene ontology analysis. Kaplan-Meir method was used to generate survival curves.

Findings

LGALS3 is highest in monocytic AML patients and those with elevated LGALS3 had significantly shorter remission duration compared to patients with lower LGALS3 levels (median 21.9 vs 51.3 weeks, p = 0.016). Pearson correlation of LGALS3 with 230 other proteins identifies a distinct set of 37 proteins positively correlated with LGALS3 expression levels with a high representation of proteins involved in AKT and ERK signaling pathways. Thirty-one proteins were negatively correlated with LGALS3 including an AKT phosphatase. Pearson correlation of proteins associated with CD74 identified 12 proteins negatively correlated with CD74 and 16 proteins that are positively correlated with CD74. CD74 network revealed strong association with CD44 signaling and a high representation of apoptosis regulators. Progeny clustering was used to build protein networks based on LGALS3 and CD74 associated proteins. A strong relationship of the LGALS3 network with the CD74 network was identified. For AML patients with both the LGALS3 and CD74 protein cluster active, median overall survival was only 24.3 weeks, median remission duration was 17.8 weeks, and no patient survived beyond one year.

Interpretation

The findings from this study identify for the first time protein networks associated with LGALS3 and CD74 in AML. Each network features unique pathway characteristics. The data also suggest that the LGALS3 network and the CD74 network each support AML cell survival and the two networks may cooperate in a novel high risk AML population.

Fund

Leukemia Lymphoma Society provided funds to SMK for RPPA study of AML patient population. Texas Leukemia provided funds to PPR and SMK to study CD74 and LGALS3 expression in AML patients using RPPA. No payment was involved in the production of this manuscript.

A quantitative analysis of heterogeneities and hallmarks in acute myelogenous leukaemia

Acute myelogenous leukaemia (AML) is associated with risk factors that are largely unknown and with a heterogeneous response to treatment. Here, we provide a comprehensive quantitative understanding of AML proteomic heterogeneities and hallmarks by using the AML proteome atlas, a proteomics database that we have newly derived from MetaGalaxy analyses, for the proteomic profiling of 205 AML patients and 111 leukaemia cell lines. The analysis of the dataset revealed 154 functional patterns based on common molecular pathways, 11 constellations of correlated functional patterns, and 13 signatures that stratify the patients’ outcomes. We find limited overlap between proteomics data and both cytogenetics and genetic mutations, and also that leukaemia cell lines show limited proteomic similarities with cells from AML patients, suggesting that a deeper focus on patient-derived samples is needed to gain disease-relevant insights. The AML proteome atlas provides a knowledge base for proteomic patterns in AML, a guide to leukaemia cell-line selection, and a broadly applicable computational approach for quantifying the heterogeneities of protein expression and proteomic hallmarks in AML.

Human papillomavirus and the landscape of secondary genetic alterations in oral cancers

Human papillomavirus (HPV) is a necessary but insufficient cause of a subset of oral squamous cell carcinomas (OSCCs) that is increasing markedly in frequency. To identify contributory, secondary genetic alterations in these cancers, we used comprehensive genomics methods to compare 149 HPV-positive and 335 HPV-negative OSCC tumor/normal pairs. Different behavioral risk factors underlying the two OSCC types were reflected in distinctive genomic mutational signatures. In HPV-positive OSCCs, the signatures of APOBEC cytosine deaminase editing, associated with anti-viral immunity, were strongly linked to overall mutational burden. In contrast, in HPV-negative OSCCs, T>C substitutions in the sequence context 5'-ATN-3' correlated with tobacco exposure. Universal expression of HPV E6*1 and E7 oncogenes was a sine qua non of HPV-positive OSCCs. Significant enrichment of somatic mutations was confirmed or newly identified in PIK3CA, KMT2D, FGFR3, FBXW7, DDX3X, PTEN, TRAF3, RB1, CYLD, RIPK4, ZNF750, EP300, CASZ1, TAF5, RBL1, IFNGR1, and NFKBIA Of these, many affect host pathways already targeted by HPV oncoproteins, including the p53 and pRB pathways, or disrupt host defenses against viral infections, including interferon (IFN) and nuclear factor kappa B signaling. Frequent copy number changes were associated with concordant changes in gene expression. Chr 11q (including CCND1) and 14q (including DICER1 and AKT1) were recurrently lost in HPV-positive OSCCs, in contrast to their gains in HPV-negative OSCCs. High-ranking variant allele fractions implicated ZNF750, PIK3CA, and EP300 mutations as candidate driver events in HPV-positive cancers. We conclude that virus-host interactions cooperatively shape the unique genetic features of these cancers, distinguishing them from their HPV-negative counterparts.

Thirty biologically interpretable clusters of transcription factors distinguish cancer type

Background

Transcription factors are essential regulators of gene expression and play critical roles in development, differentiation, and in many cancers. To carry out their regulatory programs, they must cooperate in networks and bind simultaneously to sites in promoter or enhancer regions of genes. We hypothesize that the mRNA co-expression patterns of transcription factors can be used both to learn how they cooperate in networks and to distinguish between cancer types.

Results

We recently developed a new algorithm, Thresher, that combines principal component analysis, outlier filtering, and von Mises-Fisher mixture models to cluster genes (in this case, transcription factors) based on expression, determining the optimal number of clusters in the process. We applied Thresher to the RNA-Seq expression data of 486 transcription factors from more than 10,000 samples of 33 kinds of cancer studied in The Cancer Genome Atlas (TCGA). We found that 30 clusters of transcription factors from a 29-dimensional principal component space were able to distinguish between most cancer types, and could separate tumor samples from normal controls. Moreover, each cluster of transcription factors could be either (i) linked to a tissue-specific expression pattern or (ii) associated with a fundamental biological process such as cell cycle, angiogenesis, apoptosis, or cytoskeleton. Clusters of the second type were more likely also to be associated with embryonically lethal mouse phenotypes.

Conclusions

Using our approach, we have shown that the mRNA expression patterns of transcription factors contain most of the information needed to distinguish different cancer types. The Thresher method is capable of discovering biologically interpretable clusters of genes. It can potentially be applied to other gene sets, such as signaling pathways, to decompose them into simpler, yet biologically meaningful, components.

Electronic supplementary material

The online version of this article (10.1186/s12864-018-5093-z) contains supplementary material, which is available to authorized users.

Proteogenomic Analysis of Surgically Resected Lung Adenocarcinoma

INTRODUCTION

Despite apparently complete surgical resection, approximately half of resected early-stage lung cancer patients relapse and die of their disease. Adjuvant chemotherapy reduces this risk by only 5% to 8%. Thus, there is a need for better identifying who benefits from adjuvant therapy, the drivers of relapse, and novel targets in this setting.

METHODS

RNA sequencing and liquid chromatography/liquid chromatography-mass spectrometry proteomics data were generated from 51 surgically resected non-small cell lung tumors with known recurrence status.

RESULTS

We present a rationale and framework for the incorporation of high-content RNA and protein measurements into integrative biomarkers and show the potential of this approach for predicting risk of recurrence in a group of lung adenocarcinomas. In addition, we characterize the relationship between mRNA and protein measurements in lung adenocarcinoma and show that it is outcome specific.

CONCLUSIONS

Our results suggest that mRNA and protein data possess independent biological and clinical importance, which can be leveraged to create higher-powered expression biomarkers.

Trisomy 12 chronic lymphocytic leukemia expresses a unique set of activated and targetable pathways

Although trisomy 12 (+12) chronic lymphocytic leukemia (CLL) comprises about 20% of cases, relatively little is known about its pathophysiology. These cases often demonstrate atypical morphological and immunophenotypic features, high proliferative rates, unmutated immunoglobulin heavy chain variable region genes, and a high frequency of NOTCH1 mutation. Patients with +12 CLL have an intermediate prognosis, and show higher incidences of thrombocytopenia, Richter transformation, and other secondary cancers. Despite these important differences, relatively few transcriptional profiling studies have focused on identifying dysregulated pathways that characterize +12 CLL, and most have used a hierarchical cytogenetic classification in which cases with more than one recurrent abnormality are categorized according to the abnormality with the poorest prognosis. In this study, we sought to identify protein-coding genes whose expression contributes to the unique pathophysiology of +12 CLL. To exclude the likely confounding effects of multiple cytogenetic abnormalities on gene expression, our +12 patient cohort had +12 as the sole abnormality. We profiled samples obtained from 147 treatment-naïve patients. We compared cases with +12 as the only cytogenetic abnormality to cases with only del(13q), del(11q), or diploid cytogenetics using independent discovery (n=97) and validation (n=50) sets. We demonstrate that CLL cases with +12 as the sole abnormality express a unique set of activated pathways compared to other cytogenetic subtypes. Among these pathways, we identify the NFAT signaling pathway and the immune checkpoint molecule, NT5E (CD73), which may represent new therapeutic targets.

Abstract 545: NOTCH1 co-expression analysis reveals novel insights underlying its opposing effects as an oncogenic and tumor suppressor in lung cancer

Background: Aberrations in the family of Notch receptors (1, 2, 3, and 4) have been implicated in a range of solid tumors, including lung cancer. NOTCH1 biology is complex where Notch plays an oncogenic and tumor suppressor role in lung adenocarcinoma and lung squamous cell carcinoma, respectively. Although the role of Notch in cancer development and progression has received increased appreciation in recent years, there is still a lack of understanding of the mechanisms underlying these opposing activities in lung cancer.

Methods: The Cancer Genome Atlas datasets were used to examine gene co-expression patterns of Notch1 in lung adenocarcinoma and lung squamous cell carcinoma. Biological pathways implicated by gene families were assessed using functional annotation tools (DAVID, ToppGene, and IPA). In vitro and in vivo knockdown studies assessed the functional role of Notch1.

Results: This novel co-expression analysis supports the hypothesis that NOTCH1 is co-expressed with different genes in lung adenocarcinoma and squamous cell carcinoma. Knockdown of Notch1 in vitro and in vivo support our in silico finding of opposing effects of NOTCH1. Our analysis implicates genes associated with metabolic pathways, immune pathways, angiogenesis and cell cycle that may underlie the differential role of NOTCH1 in lung adenocarcinoma and squamous cell carcinoma. In vitro and in vivo studies support our bioinformatics analysis.

Conclusion: These results demonstrate different NOTCH1 gene co-expression patterns in lung adenocarcinoma and squamous cell carcinoma. These findings provide novel insights underlying the context-dependent role of Notch as an oncogene and tumor suppressor in subtypes of lung cancer. Understanding the similarities and differences in co-expression patterns reveal novel insights, suggesting that tumor intrinsic and extrinsic pathways may underlie the dual role of NOTCH1 in lung cancer. Recognition of the mechanisms underlying NOTCH1 opposing roles in cancer may help direct development of Notch-targeted therapies as a monotherapy or in combination with approaches focusing on the tumor microenvironment.

Citation Format: Sara L. Sinicropi-Yao, David P. Carbone, Kevin R. Coombes, Joseph M. Amann, David Lopez Lopez. NOTCH1 co-expression analysis reveals novel insights underlying its opposing effects as an oncogenic and tumor suppressor in lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 545.

eIF2β, a subunit of translation-initiation factor EIF2, is a potential therapeutic target for non-small cell lung cancer

To identify novel therapeutic targets for non‐small cell lung cancer (NSCLC), we conducted an integrative study in the following 3 stages: (i) identification of potential target gene(s) through shRNA functional screens in 2 independent NSCLC cell lines; (ii) validation of the clinical relevance of identified gene(s) using public databases; and (iii) investigation of therapeutic potential of targeting the identified gene(s) in vitro. A semi‐genome‐wide shRNA screen was performed in NCI‐H358 cells, and was integrated with data from our previous screen in NCI‐H460 cells. Among genes identified in shRNA screens, 24 were present in both NCI‐H358 and NCI‐H460 cells and were considered potential targets. Among the genes, we focused on eIF2β, which is a subunit of heterotrimeric G protein EIF2 and functions as a transcription initiation factor. The eIF2β protein is highly expressed in lung cancer cell lines compared with normal bronchial epithelial cells, and gene copy number analyses revealed that eIF2β is amplified in a subset of NSCLC cell lines. Gene expression analysis using The Cancer Genome Atlas (TCGA) dataset revealed that eIF2β expression is significantly upregulated in lung cancer tissues compared with corresponding normal lung tissues. Furthermore, high eIF2β expression was correlated with poor survival in patients with lung adenocarcinoma, as shown in other cohorts using publicly available online tools. RNAi‐mediated depletion of eIF2β suppresses growth of lung cancer cells independently of p53 mutation status, in part through G1 cell cycle arrest. Our data suggest that eIF2β is a therapeutic target for lung cancer.

Decomposing the Apoptosis Pathway Into Biologically Interpretable Principal Components

Principal component analysis (PCA) is one of the most common techniques in the analysis of biological data sets, but applying PCA raises 2 challenges. First, one must determine the number of significant principal components (PCs). Second, because each PC is a linear combination of genes, it rarely has a biological interpretation. Existing methods to determine the number of PCs are either subjective or computationally extensive. We review several methods and describe a new R package, PCDimension, that implements additional methods, the most important being an algorithm that extends and automates a graphical Bayesian method. Using simulations, we compared the methods. Our newly automated procedure is competitive with the best methods when considering both accuracy and speed and is the most accurate when the number of objects is small compared with the number of attributes. We applied the method to a proteomics data set from patients with acute myeloid leukemia. Proteins in the apoptosis pathway could be explained using 6 PCs. By clustering the proteins in PC space, we were able to replace the PCs by 6 "biological components," 3 of which could be immediately interpreted from the current literature. We expect this approach combining PCA with clustering to be widely applicable.

Histone Modification Patterns Using RPPA-Based Profiling Predict Outcome in Acute Myeloid Leukemia Patients

Posttranslational histone tail modifications are known to play a role in leukemogenesis and are therapeutic targets. A global analysis of the level and patterns of expression of multiple histone-modifying proteins (HMP) in acute myeloid leukemia (AML) and the effect of different patterns of expression on outcome and prognosis has not been investigated in AML patients. Here we analyzed 20 HMP by reverse phase protein array (RPPA) in a cohort of 205 newly diagnosed AML patients. Protein levels were correlated with patient and disease characteristics, including survival and mutational state. We identified different protein clusters characterized by higher (more on) or lower (more off) expression of HMP, relative to normal CD34+ cells. On state of HMP was associated with poorer outcome compared to normal-like and a more off state. FLT3 mutated AML patients were significantly overrepresented in the more on state. DNA methylation related mutations showed no correlation with the different HMP states. In this study, we demonstrate for the first time that HMP form recurrent patterns of expression and that these significantly correlate with survival in newly diagnosed AML patients.

IntLIM: integration using linear models of metabolomics and gene expression data

BACKGROUND

Integration of transcriptomic and metabolomic data improves functional interpretation of disease-related metabolomic phenotypes, and facilitates discovery of putative metabolite biomarkers and gene targets. For this reason, these data are increasingly collected in large (> 100 participants) cohorts, thereby driving a need for the development of user-friendly and open-source methods/tools for their integration. Of note, clinical/translational studies typically provide snapshot (e.g. one time point) gene and metabolite profiles and, oftentimes, most metabolites measured are not identified. Thus, in these types of studies, pathway/network approaches that take into account the complexity of transcript-metabolite relationships may neither be applicable nor readily uncover novel relationships. With this in mind, we propose a simple linear modeling approach to capture disease-(or other phenotype) specific gene-metabolite associations, with the assumption that co-regulation patterns reflect functionally related genes and metabolites.

RESULTS

The proposed linear model, metabolite ~ gene + phenotype + gene:phenotype, specifically evaluates whether gene-metabolite relationships differ by phenotype, by testing whether the relationship in one phenotype is significantly different from the relationship in another phenotype (via a statistical interaction gene:phenotype p-value). Statistical interaction p-values for all possible gene-metabolite pairs are computed and significant pairs are then clustered by the directionality of associations (e.g. strong positive association in one phenotype, strong negative association in another phenotype). We implemented our approach as an R package, IntLIM, which includes a user-friendly R Shiny web interface, thereby making the integrative analyses accessible to non-computational experts. We applied IntLIM to two previously published datasets, collected in the NCI-60 cancer cell lines and in human breast tumor and non-tumor tissue, for which transcriptomic and metabolomic data are available. We demonstrate that IntLIM captures relevant tumor-specific gene-metabolite associations involved in known cancer-related pathways, including glutamine metabolism. Using IntLIM, we also uncover biologically relevant novel relationships that could be further tested experimentally.

CONCLUSIONS

IntLIM provides a user-friendly, reproducible framework to integrate transcriptomic and metabolomic data and help interpret metabolomic data and uncover novel gene-metabolite relationships. The IntLIM R package is publicly available in GitHub ( https://github.com/mathelab/IntLIM ) and includes a user-friendly web application, vignettes, sample data and data/code to reproduce results.

Thresher: determining the number of clusters while removing outliers

Background

Cluster analysis is the most common unsupervised method for finding hidden groups in data. Clustering presents two main challenges: (1) finding the optimal number of clusters, and (2) removing “outliers” among the objects being clustered. Few clustering algorithms currently deal directly with the outlier problem. Furthermore, existing methods for identifying the number of clusters still have some drawbacks. Thus, there is a need for a better algorithm to tackle both challenges.

Results

We present a new approach, implemented in an R package called Thresher, to cluster objects in general datasets. Thresher combines ideas from principal component analysis, outlier filtering, and von Mises-Fisher mixture models in order to select the optimal number of clusters. We performed a large Monte Carlo simulation study to compare Thresher with other methods for detecting outliers and determining the number of clusters. We found that Thresher had good sensitivity and specificity for detecting and removing outliers. We also found that Thresher is the best method for estimating the optimal number of clusters when the number of objects being clustered is smaller than the number of variables used for clustering. Finally, we applied Thresher and eleven other methods to 25 sets of breast cancer data downloaded from the Gene Expression Omnibus; only Thresher consistently estimated the number of clusters to lie in the range of 4–7 that is consistent with the literature.

Conclusions

Thresher is effective at automatically detecting and removing outliers. By thus cleaning the data, it produces better estimates of the optimal number of clusters when there are more variables than objects. When we applied Thresher to a variety of breast cancer datasets, it produced estimates that were both self-consistent and consistent with the literature. We expect Thresher to be useful for studying a wide variety of biological datasets.

Electronic supplementary material

The online version of this article (doi:10.1186/s12859-017-1998-9) contains supplementary material, which is available to authorized users.

Merging Microarray Data, Robust Feature Selection, and Predicting Prognosis in Prostate Cancer

Motivation

Individual microarray studies searching for prognostic biomarkers often have few samples and low statistical power; however, publicly accessible data sets make it possible to combine data across studies.

Method

We present a novel approach for combining microarray data across institutions and platforms. We introduce a new algorithm, robust greedy feature selection (RGFS), to select predictive genes.

Results

We combined two prostate cancer microarray data sets, confirmed the appropriateness of the approach with the Kolmogorov-Smirnov goodness-of-fit test, and built several predictive models. The best logistic regression model with stepwise forward selection used 7 genes and had a misclassification rate of 31%. Models that combined LDA with different feature selection algorithms had misclassification rates between 19% and 33%, and the sets of genes in the models varied substantially during cross-validation. When we combined RGFS with LDA, the best model used two genes and had a misclassification rate of 15%.

Availability

Affymetrix U95Av2 array data are available at http://www.broad.mit.edu/cgi-bin/cancer/datasets.cgi . The cDNA microarray data are available through the Stanford Microarray Database ( http://cmgm.stanford.edu/pbrown/ ). GeneLink software is freely available at http://bioinformatics.mdanderson.org/GeneLink/ . DNA-Chip Analyzer software is publicly available at http://biosun1.harvard.edu/complab/dchip/ .

Understanding the Characteristics of Mass Spectrometry Data through the use of Simulation

Background

Mass spectrometry is actively being used to discover disease-related proteomic patterns in complex mixtures of proteins derived from tissue samples or from easily obtained biological fluids. The potential importance of these clinical applications has made the development of better methods for processing and analyzing the data an active area of research. It is, however, difficult to determine which methods are better without knowing the true biochemical composition of the samples used in the experiments.

Methods

We developed a mathematical model based on the physics of a simple MALDI-TOF mass spectrometer with time-lag focusing. Using this model, we implemented a statistical simulation of mass spectra. We used the simulation to explore some of the basic operating characteristics of MALDI or SELDI instruments.

Results

The simulation reproduced several characteristics of actual instruments. We found that the relative mass error is affected by the time discretization of the detector (about 0.01%) and the spread of initial velocities (about 0.1%). The accuracy of calibration based on external standards decays rapidly outside the range spanned by the calibrants. Natural isotope distributions play a major role in broadening peaks associated with individual proteins. The area of a peak is a more accurate measure of its size than the height.

Conclusions

The model described here is capable of simulating realistic mass spectra. The simulation should become a useful tool for generating spectra where the true inputs are known, allowing researchers to evaluate the performance of new methods for processing and analyzing mass spectra.

Availability

http://bioinformatics.mdanderson.org/cromwell.html

Bias, Randomization, and Ovarian Proteomic Data: A Reply to “Producers and Consumers”

Proteomic patterns derived from mass spectrometry have recently been put forth as potential biomarkers for the early diagnosis of cancer. This approach has generated much excitement, particularly as initial results reported on SELDI profiling of serum suggested that near perfect sensitivity and specificity could be achieved in diagnosing ovarian cancer. However, more recent reports have suggested that much of the observed structure could be due to the presence of experimental bias. A rebuttal to the findings of bias, subtitled “Producers and Consumers”, lists several objections. In this paper, we attempt to address these objections. While we continue to find evidence of experimental bias, we emphasize that the problems found are associated with experimental design and processing, and can be avoided in future studies.

Identification of proteasomal catalytic subunit PSMA6 as a therapeutic target for lung cancer

To identify potential therapeutic targets for lung cancer, we performed semi‐genome‐wide shRNA screening combined with the utilization of genome‐wide expression and copy number data. shRNA screening targeting 5043 genes in NCI‐H460 identified 51 genes as candidates. Pathway analysis revealed that the 51 genes were enriched for the five pathways, including ribosome, proteasome, RNA polymerase, pyrimidine metabolism and spliceosome pathways. We focused on the proteasome pathway that involved six candidate genes because its activation has been demonstrated in diverse human malignancies, including lung cancer. Microarray expression and array CGH data showed that PSMA6, a proteasomal subunit of a 20S catalytic core complex, was highly expressed in lung cancer cell lines, with recurrent gene amplifications in some cases. Therefore, we further examined the roles of PSMA6 in lung cancer. Silencing of PSMA6 induced apoptosis or G2/M cell cycle arrest in cancer cell lines but not in an immortalized normal lung cell line. These results suggested that PSMA6 serves as an attractive target with a high therapeutic index for lung cancer.

Risk Factors of 131I-Induced Salivary Gland Damage in Thyroid Cancer Patients

CONTEXT

Sialadenitis and xerostomia are major adverse effects of ¹³¹I therapy in thyroid cancer patients. The risk factors for these adverse effects, other than administered activity of ¹³¹I, have not been investigated.

OBJECTIVE

The aim of this study is to identify risk factors for ¹³¹I-induced salivary gland damage among follicular cell-derived thyroid cancer patients.

DESIGN

We enrolled 216 thyroid cancer patients who visited The Ohio State University Wexner Medical Center between April 2013 and April 2014. Symptoms of xerostomia and sialadenitis were identified via questionnaire and medical record search. To validate the findings in a large cohort, we retrospectively searched for ICD-9/10 codes for sialadenitis, xerostomia, and autoimmune disease associated with Sjögren's syndrome (AID-SS) in our existing database (n = 1507). Demographic and clinical information was extracted from medical records. Multivariate analyses were performed to identify independent predictors for salivary gland damage.

RESULTS

¹³¹I treatment associated with higher incidence of xerostomia and sialadenitis. Patients with xerostomia had 46 mCi higher mean cumulative ¹³¹I activity and 21 mCi higher mean first-administered ¹³¹I activity than patients without xerostomia. Increased age associated with higher incidence of xerostomia, and females had a higher incidence of sialadenitis. Patients who experienced sialadenitis before ¹³¹I therapy had higher sialadenitis incidence after ¹³¹I therapy. ¹³¹I-treated patients diagnosed with AID-SS, whether before or after ¹³¹I treatment, had a higher incidence of xerostomia and sialadenitis among ¹³¹I-treated patients.

CONCLUSION

Risk factors for ¹³¹I-induced salivary gland damage include administered ¹³¹I activity, age, gender, history of sialadenitis before ¹³¹I treatment, and AID-SS diagnosis.

The BATTLE-2 Study: A Biomarker-Integrated Targeted Therapy Study in Previously Treated Patients With Advanced Non-Small-Cell Lung Cancer

PURPOSE

By applying the principles of real-time biopsy, biomarker-based, adaptively randomized studies in non-small-cell lung cancer (NSCLC) established by the Biomarker-Integrated Approaches of Targeted Therapy for Lung Cancer Elimination (BATTLE) trial, we conducted BATTLE-2 (BATTLE-2 Program: A Biomarker-Integrated Targeted Therapy Study in Previously Treated Patients With Advanced Non-Small Cell Lung Cancer), an umbrella study to evaluate the effects of targeted therapies focusing on KRAS-mutated cancers.

PATIENTS AND METHODS

Patients with advanced NSCLC (excluding sensitizing EGFR mutations and ALK gene fusions) refractory to more than one prior therapy were randomly assigned, stratified by KRAS status, to four arms: (1) erlotinib, (2) erlotinib plus MK-2206, (3) MK-2206 plus AZD6244, or (4) sorafenib. Tumor gene expression profiling-targeted next-generation sequencing was performed to evaluate predictive and prognostic biomarkers.

RESULTS

Two hundred patients, 27% with KRAS-mutated (KRAS mut+) tumors, were adaptively randomly assigned to erlotinib (n = 22), erlotinib plus MK-2206 (n = 42), MK-2206 plus AZD6244 (n = 75), or sorafenib (n = 61). In all, 186 patients were evaluable, and the primary end point of an 8-week disease control rate (DCR) was 48% (arm 1, 32%; arm 2, 50%; arm 3, 53%; and arm 4, 46%). For KRAS mut+ patients, DCR was 20%, 25%, 62%, and 44% whereas for KRAS wild-type patients, DCR was 36%, 57%, 49%, and 47% for arms 1, 2, 3, and 4, respectively. Median progression-free survival was 2.0 months, not different by KRAS status, 1.8 months for arm 1, and 2.5 months for arms 2 versus arms 3 and 4 in KRAS mut+ patients (P = .04). Median overall survival was 6.5 months, 9.0 and 5.1 months for arms 1 and 2 versus arms 3 and 4 in KRAS wild-type patients (P = .03). Median overall survival was 7.5 months in mesenchymal versus 5 months in epithelial tumors (P = .02).

CONCLUSION

Despite improved progression-free survival on therapy that did not contain erlotinib for KRAS mut+ patients and improved prognosis for mesenchymal tumors, better biomarker-driven treatment strategies are still needed.

Low expression of ASH2L protein correlates with a favorable outcome in acute myeloid leukemia

ASH2L encodes a trithorax group protein that is a core component of all characterized mammalian histone H3K4 methyltransferase complexes, including mixed lineage leukemia (MLL) complexes. ASH2L protein levels in primary leukemia patient samples have not yet been defined. We analyzed ASH2L protein expression in 511 primary AML patient samples using reverse phase protein array (RPPA) technology. We discovered that ASH2L expression is significantly increased in a subset of patients carrying fms-related tyrosine kinase 3 (FLT3) mutations. Furthermore, we observed that low levels of ASH2L are associated with increased overall survival. We also compared ASH2L levels to the expression of 230 proteins previously analyzed on this array. ASH2L expression was inversely correlated with 32 proteins, mostly involved in cell adhesion and cell cycle inhibition, while a positive correlation was observed for 50 proteins, many of which promote cell proliferation. Together, these results indicate that a lower level of ASH2L protein is beneficial to AML patients.

Epidermal growth factor receptor is expressed and active in a subset of acute myeloid leukemia

The epidermal growth factor receptor (EGFR) inhibitor erlotinib has been shown to induce complete remission of acute myeloid leukemia (AML) in two patients with concurrent lung cancer and raised attention for a role of EGFR in AML whereas a recent phase II clinical study with gefitinib in AML demonstrated a negative result on the outcome. However, from several studies, EGFR expression in AML is poorly defined and the role of EGFR in AML remains unclear. Herein, we report the results of EGFR expression in AML of large cohorts of adult and pediatric AML patients with the data of total protein and phosphorylation levels of EGFR. Our data conclude that there is the expression of EGFR at the protein level in a subset of AML, which was identified to be functionally active in ~15 % of AML patients. This suggests that future studies need to be conducted with a subset of AML patients characterized by high EGFR expression.

Selecting Reliable mRNA Expression Measurements Across Platforms Improves Downstream Analysis

With increasing use of publicly available gene expression data sets, the quality of the expression data is a critical issue for downstream analysis, gene signature development, and cross-validation of data sets. Thus, identifying reliable expression measurements by leveraging multiple mRNA expression platforms is an important analytical task. In this study, we propose a statistical framework for selecting reliable measurements between platforms by modeling the correlations of mRNA expression levels using a beta-mixture model. The model-based selection provides an effective and objective way to separate good probes from probes with low quality, thereby improving the efficiency and accuracy of the analysis. The proposed method can be used to compare two microarray technologies or microarray and RNA sequencing measurements. We tested the approach in two matched profiling data sets, using microarray gene expression measurements from the same samples profiled on both Affymetrix and Illumina platforms. We also applied the algorithm to mRNA expression data to compare Affymetrix microarray data with RNA sequencing measurements. The algorithm successfully identified probes/genes with reliable measurements. Removing the unreliable measurements resulted in significant improvements for gene signature development and functional annotations.

MLL-Rearranged Acute Lymphoblastic Leukemias Activate BCL-2 through H3K79 Methylation and Are Sensitive to the BCL-2-Specific Antagonist ABT-199

Targeted therapies designed to exploit specific molecular pathways in aggressive cancers are an exciting area of current research. Mixed Lineage Leukemia (MLL) mutations such as the t(4;11) translocation cause aggressive leukemias that are refractory to conventional treatment. The t(4;11) translocation produces an MLL/AF4 fusion protein that activates key target genes through both epigenetic and transcriptional elongation mechanisms. In this study, we show that t(4;11) patient cells express high levels of BCL-2 and are highly sensitive to treatment with the BCL-2-specific BH3 mimetic ABT-199. We demonstrate that MLL/AF4 specifically upregulates the BCL-2 gene but not other BCL-2 family members via DOT1L-mediated H3K79me2/3. We use this information to show that a t(4;11) cell line is sensitive to a combination of ABT-199 and DOT1L inhibitors. In addition, ABT-199 synergizes with standard induction-type therapy in a xenotransplant model, advocating for the introduction of ABT-199 into therapeutic regimens for MLL-rearranged leukemias.

A Patient-Derived, Pan-Cancer EMT Signature Identifies Global Molecular Alterations and Immune Target Enrichment Following Epithelial-to-Mesenchymal Transition

PURPOSE

We previously demonstrated the association between epithelial-to-mesenchymal transition (EMT) and drug response in lung cancer using an EMT signature derived in cancer cell lines. Given the contribution of tumor microenvironments to EMT, we extended our investigation of EMT to patient tumors from 11 cancer types to develop a pan-cancer EMT signature.

EXPERIMENTAL DESIGN

Using the pan-cancer EMT signature, we conducted an integrated, global analysis of genomic and proteomic profiles associated with EMT across 1,934 tumors including breast, lung, colon, ovarian, and bladder cancers. Differences in outcome and in vitro drug response corresponding to expression of the pan-cancer EMT signature were also investigated.

RESULTS

Compared with the lung cancer EMT signature, the patient-derived, pan-cancer EMT signature encompasses a set of core EMT genes that correlate even more strongly with known EMT markers across diverse tumor types and identifies differences in drug sensitivity and global molecular alterations at the DNA, RNA, and protein levels. Among those changes associated with EMT, pathway analysis revealed a strong correlation between EMT and immune activation. Further supervised analysis demonstrated high expression of immune checkpoints and other druggable immune targets, such as PD1, PD-L1, CTLA4, OX40L, and PD-L2, in tumors with the most mesenchymal EMT scores. Elevated PD-L1 protein expression in mesenchymal tumors was confirmed by IHC in an independent lung cancer cohort.

CONCLUSIONS

This new signature provides a novel, patient-based, histology-independent tool for the investigation of EMT and offers insights into potential novel therapeutic targets for mesenchymal tumors, independent of cancer type, including immune checkpoints.