03/05/08: Three ENAR practice talks

This student seminar will be a collection of ENAR practice talks from JoAnna Scott, Rebecca Hubbard, and Leslie Taylor. Each talk will be approximately 15 minutes long.

JoAnna Scott

Title: Vaccine Efficacy Trials using Stepped Wedge Design

Abstract: There have been many changes in the way that the effectiveness of a vaccine has been evaluated, in particular for HIV vaccine trials. Of note have been the Test of Concept or Phase IIB trials, seen recently in the STEP 502 and 503 trials. These designs leads to some interesting questions regarding what the next step should be if these trials show were to show efficacy. The Stepped Wedge design is a useful trial design that can be very effective in addressing the issues that would occur if the Phase IIB trials were efficacious. In the Stepped Wedge design, time when an intervention is introduced is randomized and all clusters will eventually receive the intervention. For this talk, I will discuss how the Stepped Wedge design is a useful one for vaccine efficacy trials and I also will discuss different methods for estimating and testing vaccine efficacy within a Stepped Wedge cluster randomized trial.

Rebecca Hubbard

Modeling risk factors for Alzheimer’s disease progression using a
non-homogeneous Markov process

Identifying individuals at high risk of developing Alzheimer’s
disease (AD) is important for understanding the natural history of the disease
and effectively targeting interventions. Subjects suffering from mild cognitive
impairment (MCI) are one group with high probability of progression to AD.
Estimating rate of progression from normal cognition to MCI and AD is
challenging because cognitive status is ascertained only at irregular follow-up
times giving rise to interval censored data.

Moreover, progression rates are known to be non-constant with respect to age.
Markov process models are useful for characterizing transition rates in multi-state disease processes with interval censoring. However, limited methods exist for temporally non-homogeneous multi-state processes. We propose a non-homogeneous Markov process model to characterize transitions between disease states defined by normal cognition, MCI, AD, and death. Risk factors for increased rates of transition are introduced via a regression model for
elements of the baseline transition intensity matrix. We apply this model to a longitudinal study of subjects evaluated at the Alzheimer s Disease Centers.

Leslie Taylor

Dealing with noncompliance and nonresponse in a clustered
encouragement design study.

Well-designed randomized clinical trials are a powerful tool for
investigating causal treatment effects, but in human trials there are
oftentimes problems of noncompliance. This is particularly a problem in
encouragement design studies, where encouragement to take the treatment, rather
than the treatment itself, is randomized. We consider a ‘clustered
encouragement design’, meaning that the randomization is at the level of the
clusters (e.g. physicians), but the compliance with assignment is at the level
of the units (e.g. patients) within clusters (Frangakis et al. 2002).
Furthermore, there is a problem of outcome nonresponse, as is typical in most
clinical trials. Frangakis et al. (2002) proposed a Bayesian methodology for
causal inference in a clustered encouragement design setting. We extend their
setting to one in which there is outcome nonresponse, and we propose an
alternative approach to

02/27/08: Time-Dependent Receiver Operating Characteristic Curves for Early Markers of Event Time Outcomes

Yuying Jin

The receiver operating characteristic (ROC) curve is a popular diagnostic tool assessing the diagnostic accuracy of a putative biomarker. When a biomarker is used for prognosis, often the marker is not measured concurrently with disease outcome. Further, the disease onset time can be censored. Therefore the classic concept of ROC curves needs to be extended to integrate both the time-varying nature of the marker and the clinical onset time of the disease. Naive estimates of predictive accuracy ignore the censoring can be biased. Incorporating the time dimension in ROC curve analysis has recently been an area of active research. We contrast the forms and attributes of various time-dependent ROC estimators discussed in the literature. The bias and efficiency of these estimators are investigated in simulation studies. Recommendations for their uses in practical situations are given.

02/20/08: HAART and the Heart: Treating Protease Inhibitor Dyslipidemia in HIV+ Patients

Betsy Teeple

With the introduction of protease inhibitor (PI) based highly-active antiretroviral therapy (HAART) into clinical practice in 1996, HIV infection has transformed into a manageable chronic disease. However, by 1997 and 1998, adverse metabolic effects were recognized as toxicities associated with long term PI use. These complications include increased triglycerides, increased LDL and VLDL cholesterol, and decreased HDL cholesterol levels, all of which have been associated with increased risk of coronary heart disease risk in the HIV negative population. In my biology project, I will discuss HIV treatments, including PIs, the lipid metabolism pathway, which lipid metabolism mechanisms are disrupted by PIs, dyslipidemia treatments and their pathways in the HIV negative population, and the
drug-drug interaction difficulties behind treating both HIV and PI-associated lipid metabolism disorders at the same time.

02/13/08: Statistical modeling for protein identification using mass spectrometry

Qunhua Li

Protein identification using mass spectrometry is a high-throughput
way to identify proteins in biological samples. In this talk, we use
statistical approaches to model two key steps in this process, namely,
identifying peptides from mass spectra using protein database search
and identifying proteins from putative peptide identifications. Both
problems are featured by high-dimensional data and low signal-to-noise
ratio.

For the problem of peptide identification, we developed a
likelihood-based algorithm based on a latent variable model, which
measures the likelihood that the observed spectrum arises from the
theoretical spectra predicted from each peptide contained in a protein
database. By carefully modeling the noise structure, our probability
model takes account of multiple sources of noise in the data and
extract some of the subtle signals which other methods miss. In
addition, our likelihood-based approach also provides natural measures
for assessing the uncerteinty of each identification.

The task of protein identification essentially is to assess the
evidence of presence for proteins constructed from putative peptides
identifications. We develop an unsupervised protein identification
algorithm based on a nested mixture model, which incorporates the
evidence feedback between peptide level and protein level. Our model
essentially is a model-based clustering method, which jointly
validates the correctness of peptide identification and infers the
evidence of presence of proteins by simultaneously clustering the
labels of peptides and proteins. Using a yeast dataset, we show that
our method has a competitive performance over leading products on
protein identification and peptide validation.

Change of venue for 02/06/08

Today the carpet is being laid in the reception area and hallway in front of the conference room.  There will be no access to both the kitchen area and the F-Wing Conference Room.

The location of today’s Student Seminar has been moved to the Hlth Svcs Conference Room, H-670.