A5101715307- Statistical Methods in Clinical Trials
- Cancer Genomics and Diagnostics
- Gene expression and cancer classification
- Bioinformatics and Genomic Networks
- Statistical Methods and Inference
- Health Systems, Economic Evaluations, Quality of Life
- Optimal Experimental Design Methods
- Molecular Biology Techniques and Applications
- Lung Cancer Treatments and Mutations
- Advanced Causal Inference Techniques
- Lymphoma Diagnosis and Treatment
- Genetic factors in colorectal cancer
- Immunotherapy and Immune Responses
- Colorectal Cancer Treatments and Studies
- vaccines and immunoinformatics approaches
- PARP inhibition in cancer therapy
- Cancer Immunotherapy and Biomarkers
- Statistical Methods and Bayesian Inference
- Meta-analysis and systematic reviews
- Ovarian cancer diagnosis and treatment
- Quantum Mechanics and Applications
- Monoclonal and Polyclonal Antibodies Research
- Computational Drug Discovery Methods
- HER2/EGFR in Cancer Research
- Prostate Cancer Treatment and Research
Lismore Base Hospital
2024
National Cancer Institute
2012-2023
New York City Police Department
2023
Rochester Institute of Technology
2023
Lawrence Berkeley National Laboratory
2020
Cleveland Clinic
2020
Software (Spain)
2018
National Institutes of Health
2008-2017
University of Maryland, Baltimore
2016
University of Alabama at Birmingham
2016
We describe the use of cubic splines in regression models to represent relationship between response variable and a vector covariates. This simple method can help prevent problems that result from inappropriate linearity assumptions. compare restricted spline non-parametric procedures for characterizing age survival Stanford Heart Transplant data. also provide an illustrative example cancer therapeutics.
In controlled clinical trials there are usually several prognostic factors known or thought to influence the patient's ability respond treatment. Therefore, method of sequential treatment assignment needs be designed so that balance is simultaneously achieved across all such patients factor. Traditional methods restricted randomization as "permuted blocks within strata" prove inadequate once number strata, combinations factor levels, approaches sample size. A new general procedure for...
Abstract Background Cross-validation (CV) is an effective method for estimating the prediction error of a classifier. Some recent articles have proposed methods optimizing classifiers by choosing classifier parameter values that minimize CV estimate. We evaluated validity using estimate optimized as true expected on independent data. Results used to optimize classification parameters two kinds classifiers; Shrunken Centroids and Support Vector Machines (SVM). Random training datasets were...
Targeted magnetic resonance (MR)/ultrasound fusion prostate biopsy has been shown to detect cancer. The implications of targeted alone vs standard extended-sextant or the 2 modalities combined are not well understood.To assess and approaches for diagnosis intermediate- high-risk cancer.Prospective cohort study 1003 men undergoing both concurrently from 2007 through 2014 at National Cancer Institute in United States. Patients were referred elevated level prostate-specific antigen (PSA)...
In genomic studies, thousands of features are collected on relatively few samples. One the goals these studies is to build classifiers predict outcome future observations. There three inherent steps this process: feature selection, model selection and prediction assessment. With a focus assessment, we compare several methods for estimating 'true' error in presence selection.For small where selected from candidates, resubstitution simple split-sample estimates seriously biased. samples,...
Using current diagnostic criteria, primary mediastinal B cell lymphoma (PMBL) cannot be distinguished from other types of diffuse large (DLBCL) reliably. We used gene expression profiling to develop a more precise molecular diagnosis PMBL. PMBL patients were considerably younger than DLBCL patients, and their lymphomas frequently involved thoracic structures but not extrathoracic sites typical DLBCLs. had relatively favorable clinical outcome, with 5-yr survival rate 64% compared 46% for...
The development of tumor biomarkers ready for clinical use is complex. We propose a refined system biomarker study design, conduct, analysis, and evaluation that incorporates hierarchal level evidence scale marker studies, including those using archived specimens. Although fully prospective randomized trials to evaluate the medical utility prognostic or predictive are gold standard, such costly, so we discuss more efficient indirect "prospective–retrospective" designs In particular, new...
The National Cancer Institute (NCI) is implementing a large-scale in vitro drug-screening program that requires very efficient automated assay of drug effects on tumor cell viability or growth. Many laboratories worldwide have adopted microculture based metabolic reduction 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT). However, because certain technical advantages to use the protein-binding dye sulforhodamine B (SRB) screening application, detailed comparison data...
The distinction between Burkitt's lymphoma and diffuse large-B-cell is crucial because these two types of require different treatments. We examined whether gene-expression profiling could reliably distinguish from lymphoma.Tumor-biopsy specimens 303 patients with aggressive lymphomas were profiled for gene expression also classified according to morphology, immunohistochemistry, detection the t(8;14) c-myc translocation.A classifier based on correctly identified all 25 pathologically...
Microarray techniques provide a valuable way of characterizing the molecular nature disease. Unfortunately expense and limited specimen availability often lead to studies with small sample sizes. This makes accurate estimation variability difficult, since variance estimates made on gene by basis will have few degrees freedom, assumption that all genes share equal is unlikely be true.We propose model which within variances are drawn from an inverse gamma distribution, whose parameters...
Evaluation of evidence that treatment efficacy varies substantially among different subsets patients is an important feature the analysis large clinical trials. Qualitative or crossover interactions are said to occur when one superior for some and alternative other subsets. A non-crossover interaction arises there variation in magnitude, but not direction, effects Some authors use term quantitative mean interaction. Non-crossover usually less importance than qualitative interactions, which...
Background: Many cancer patients in phase I clinical trials are treated at doses of chemotherapeutic agents that below the biologically active level, thus reducing their chances for therapeutic benefit. Current often take a long time to complete and provide little information about interpatient variability or cumulative toxicity. Purpose: Our objective was develop alternative designs so fewer subtherapeutic dose levels, reduced duration, important (i.e., toxicity maximum tolerated dose)...
Human autoimmune diseases are thought to develop through a complex combination of genetic and environmental factors. Genome-wide linkage searches inflammatory/immune disorders have identified large number non-major histocompatibility loci that collectively contribute disease susceptibility. A comparison was made the results from 23 published or immune-mediated genome-wide scans. included multiple sclerosis, Crohn's disease, familial psoriasis, asthma, type-I diabetes (IDDM). Experimental...
BRB-ArrayTools is an integrated software system for the comprehensive analysis of DNA microarray experiments. It was developed by professional biostatisticians experienced in design and studies incorporates methods leading statistical laboratories. The designed use biomedical scientists who wish to have access state-of-the-art gene expression data receive training high dimensional data. provides most extensive set tools available predictive classifier development complete cross-validation....
In spite of the controversy over role randomized clinical trials in medical research, rationale underlying such remains persuasive as compared to recent suggestions for alternative non-randomized studies those relying on use historical controls and adjustment technics. Others have suggested that statistical innovations improving trials, including adaptive allocation treatment patients sequential stopping procedures, are underutilized. These innovations, though theoretically interesting, not...
Abstract Neural networks have received considerable attention recently, mostly by non‐statisticians. They are considered many to be very promising tools for classification and prediction. In this paper we present an approach modelling censored survival data using the input—output relationship associated with a simple feed‐forward neural network as basis non‐linear proportional hazards model. This can extended other models used data. The parameters estimated method of maximum likelihood....