Anthropoid humeri from the late Eocene of Egypt

A number of recent studies have, by necessity, placed a great deal of emphasis on the dental evidence for Paleogene anthropoid interrelationships, but cladistic analyses of these data have led to the erection of phylogenetic hypotheses that appear to be at odds with biogeographic and stratigraphic considerations. Additional morphological data from the cranium and postcranium of certain poorly understood Paleogene primates are clearly needed to help test whether such hypotheses are tenable. Here we describe humeri attributable to Proteopithecus sylviae and Catopithecus browni, two anthropoids from late Eocene sediments of the Fayum Depression in Egypt. Qualitative and morphometric analyses of these elements indicate that humeri of the oligopithecine Catopithecus are more similar to early Oligocene propliopithecines than they are to any other Paleogene anthropoid taxon, and that Proteopithecus exhibits humeral similarities to parapithecids that may be symplesiomorphies of extant (or “crown”) Anthropoidea. The humeral morphology of Catopithecus is consistent with certain narrowly distributed dental apomorphies—such as the loss of the upper and lower second premolar and the development of a honing blade for the upper canine on the lower third premolar—which suggest that oligopithecines constitute the sister group of a clade containing propliopithecines and Miocene-Recent catarrhines and are not most closely related to Proteopithecus as has recently been proposed.

Prediction of protein folding class using global description of amino acid sequence.

We present a method for predicting protein folding class based on global protein chain description and a voting process. Selection of the best descriptors was achieved by a computer-simulated neural network trained on a data base consisting of 83 folding classes. Protein-chain descriptors include overall composition, transition, and distribution of amino acid attributes, such as relative hydrophobicity, predicted secondary structure, and predicted solvent exposure. Cross-validation testing was performed on 15 of the largest classes. The test shows that proteins were assigned to the correct class (correct positive prediction) with an average accuracy of 71.7%, whereas the inverse prediction of proteins as not belonging to a particular class (correct negative prediction) was 90-95% accurate. When tested on 254 structures used in this study, the top two predictions contained the correct class in 91% of the cases.