Parallel computing in biomedical research.

Scalable parallel computer architectures provide the computational performance needed for advanced biomedical computing problems. The National Institutes of Health have developed a number of parallel algorithms and techniques useful in determining biological structure and function. These applications include processing electron micrographs to determine the three-dimensional structure of viruses, calculating the solvent-accessible surface area of proteins to help predict the three-dimensional conformation of these molecules from their primary structures, and searching for homologous DNA or amino acid sequences in large biological databases. Timing results demonstrate substantial performance improvements with parallel implementations compared with conventional sequential systems.

Childhood-onset schizophrenia/autistic disorder and t(1;7) reciprocal translocation: identification of a BAC contig spanning the translocation breakpoint at 7q21.

Childhood-onset schizophrenia (COS) is defined by the development of first psychotic symptoms by age 12. While recruiting patients with COS refractory to conventional treatments for a trial of atypical antipsychotic drugs, we discovered a unique case who has a familial t(1;7)(p22;q21) reciprocal translocation and onset of psychosis at age 9. The patient also has symptoms of autistic disorder, which are usually transient before the first psychotic episode among 40-50% of the childhood schizophrenics but has persisted in him even after the remission of psychosis. Cosegregating with the translocation, among the carriers in the family available for the study, are other significant psychopathologies, including alcohol/drug abuse, severe impulsivity, and paranoid personality and language delay. This case may provide a model for understanding the genetic basis of schizophrenia or autism. Here we report the progress toward characterization of genomic organization across the translocation breakpoint at 7q21. The polymorphic markers, D7S630/D7S492 and D7S2410/D7S646, immediately flanking the breakpoint, may be useful for further confirming the genetic linkage for schizophrenia or autism in this region. Am. J. Med. Genet. (Neuropsychiatr. Genet.) 96:749-753, 2000. Published 2000 Wiley-Liss, Inc.

Precision and accuracy of regional radioactivity quantitation using the maximum likelihood EM reconstruction algorithm.

The imaging characteristics of maximum likelihood (ML) reconstruction using the EM algorithm for emission tomography have been extensively evaluated. There has been less study of the precision and accuracy of ML estimates of regional radioactivity concentration. The authors developed a realistic brain slice simulation by segmenting a normal subject's MRI scan into gray matter, white matter, and CSF and produced PET sinogram data with a model that included detector resolution and efficiencies, attenuation, scatter, and randoms. Noisy realizations at different count levels were created, and ML and filtered backprojection (FBP) reconstructions were performed. The bias and variability of ROI values were determined. In addition, the effects of ML pixel size, image smoothing and region size reduction were assessed. Hit estimates at 3,000 iterations (0.6 sec per iteration on a parallel computer) for 1-cm(2) gray matter ROIs showed negative biases of 6%+/-2% which can be reduced to 0%+/-3% by removing the outer 1-mm rim of each ROI. FBP applied to the full-size ROIs had 15%+/-4% negative bias with 50% less noise than hit. Shrinking the FBP regions provided partial bias compensation with noise increases to levels similar to ML. Smoothing of ML images produced biases comparable to FBP with slightly less noise. Because of its heavy computational requirements, the ML algorithm will be most useful for applications in which achieving minimum bias is important.