Mendel-GFDb and Mendel-ESTS: databases of plant gene families and ESTs annotated with gene family numbers and gene family names.

There is no control over the information provided with sequences when they are deposited in the sequence databases. Consequently mistakes can seed the incorrect annotation of other sequences. Grouping genes into families and applying controlled annotation overcomes the problems of incorrect annotation associated with individual sequences. Two databases (http://www.mendel.ac.uk) were created to apply controlled annotation to plant genes and plant ESTs: Mendel-GFDb is a database of plant protein (gene) families based on gapped-BLAST analysis of all sequences in the SWISS-PROT family of databases. Sequences are aligned (ClustalW) and identical and similar residues shaded. The families are visually curated to ensure that one or more criteria, for example overall relatedness and/or domain similarity relate all sequences within a family. Sequence families are assigned a 'Gene Family Number' and a unified description is developed which best describes the family and its members. If authority exists the gene family is assigned a 'Gene Family Name'. This information is placed in Mendel-GFDb. Mendel-ESTS is primarily a database of plant ESTs, which have been compared to Mendel-GFDb, completely sequenced genomes and domain databases. This approach associated ESTs with individual sequences and the controlled annotation of gene families and protein domains; the information being placed in Mendel-ESTS. The controlled annotation applied to genes and ESTs provides a basis from which a plant transcription database can be developed.

Myocardial suppression in vivo by spin locking with composite pulses.

Improved myocardium-blood contrast by myocardial suppression resulting from T1 rho-weighting in contrast-enhanced, gradient-echo, bright-blood cine images, acquired at 1.5T, is shown. In the standard images, blood has twice the intensity of muscle. In similar T1 rho-weighted images, it has 3-4 times the intensity of muscle. A composite spin-lock pulse before each observation pulse provides T1 rho-weighting. A typical pulse was: 90y-135x-360x-135x-90(-y) with element durations: 0.84, 1.26, 8.12, 1.26, and 0.84 ms. The tolerance of this composite pulse to shimmering and frequency errors allows spin locking with comparatively weak RF and therefore low specific absorption rate (SAR). Initial clinical evaluation on patients with poor ventricular function demonstrates both a qualitative and quantitative improvement in delineation of myocardial borders.