Fred Hutchinson Cancer Research Center

About: Fred Hutchinson Cancer Research Center is a nonprofit organization based out in Cape Town, South Africa. It is known for research contribution in the topics: Population & Transplantation. The organization has 12322 authors who have published 30954 publications receiving 2288772 citations. The organization is also known as: Fred Hutch & The Hutch.

Epidemiology of Aspergillus Infections in a Large Cohort of Patients Undergoing Bone Marrow Transplantation

Abstract: To investigate the incidence, risk factors, and outcome of Aspergillus infections among marrow transplant recipients, records from 2496 patients were reviewed, and 214 patients had Aspergillus organisms identified. Of these, 158 had invasive aspergillosis, 44 were colonized, and 12 had contaminated cultures. The incidence of invasive aspergillosis increased from 5.7% to 11.2% during the study. The onset of infection was bimodal, peaking 16 and 96 days after transplant. For patients within 40 days after transplant, underlying disease, donor type, season, and transplant outside of laminar air flow rooms were associated with significant risk for invasive aspergillosis. For patients >40 days after transplant, age, underlying disease, donor type, graft-versus-host disease, neutropenia, and corticosteroid use were associated with increased risk of aspergillosis. Only 31% of infected patients were neutropenic at the time of diagnosis. The risk factors for aspergillosis depend on the time after marrow transplant and include both host and environmental characteristics.

The joint graphical lasso for inverse covariance estimation across multiple classes

Abstract: We consider the problem of estimating multiple related Gaussian graphical models from a high-dimensional data set with observations belonging to distinct classes. We propose the joint graphical lasso, which borrows strength across the classes in order to estimate multiple graphical models that share certain characteristics, such as the locations or weights of nonzero edges. Our approach is based upon maximizing a penalized log likelihood. We employ generalized fused lasso or group lasso penalties, and implement a fast ADMM algorithm to solve the corresponding convex optimization problems. The performance of the proposed method is illustrated through simulated and real data examples.