Tid: 2 december 2002 kl 1515-1600
Plats : Seminarierummet 3733, Institutionen för matematik, KTH, Lindstedts väg 25, plan 7. Karta!
Föredragshållare: Ronny Alvang.
Titel: Classification of protein crystal unit cells into visually distinct types by Principal Components Analysis.
Sammanfattning: Software, written in Python, for the multivariate statistical method Principal Components Analysis (PCA) was developed, and successfully used in a test case to study extremely noisy electron microscopic images of the crystal cells of a sheet of crystallized protein: Microsomal Glutathione Transferase. Only in the average of many of these extremely noisy pictures can any structure be seen. In this average there are seemingly two different protein unit cells: a hexamer (with six-fold symmetry) and a trimer. It was believed that the hexamer is really an average of trimer images in two different orientations. I managed to show, with the capacity of PCA to separate different classes of data, that the hexamer is indeed composed of trimers, but in four, not two, orientations. This result should be verified on more crystals.