Genomic Profiling of Blood for Stroke Diagnosis
George Mason University, 2009 - Cerebrovascular disease - 172 pages
Genomic profiling of blood for stroke diagnosis is avidly being researched today. For the reason, current tools available to diagnose stroke (i.e., Computed Tomography Scan, Magnetic Resonance Imaging) are not universally available. In addition, physician inexperience and patient acuity deter their use even when available, rendering delayed diagnosis, misdiagnosis, or no diagnosis at all. However, research performed thus far has not asked nor answered whether common genomic markers for stroke event (i.e., ischemic or hemorrhagic) and discriminative genomic markers for stroke type (i.e., ischemic vs. hemorrhagic) exist in human blood post stroke. Both types of markers are critically necessary for a successful stroke diagnosis. This dissertation represents the first discovery work in this regard. Specifically, blood-borne genomic markers for stroke event and stroke type were screened, identified, and used in the construction of diagnostic models for stroke event and stroke type. Results provide unwavering evidence that many non-stroke conditions affect marker expression similar to stroke; complicating the ability to discern stroke from non-stroke conditions. Reference-based modeling was required to circumvent this complication; providing for models having a combined cross-validated accuracy rate of ∼91% for stroke event and ∼77% for stroke type. While, investigation of marker biology provided an astounding and novel postulation of the underlying mechanism of this action. That is, stroke appears to arise in consequence to a persistent infection by Francisella tularensis, a subspecies of the same genus, or a new uncharacterized genus of bacteria having similar characteristics to Francisella tularensis. In fact, the occurrence of stroke, frequency of stroke, and type of stroke incurred appears to be a function of time infected, mean arterial blood pressure, amount of circulating iron, amount of circulating fat, personal genome, personal expressome, and personal habits.
What people are saying - Write a review
We haven't found any reviews in the usual places.