741 proteins were associated with DNA-binding folds. Of the remaining 1121 ORFs that could not be modeled at this time, 600 are annotated as conserved hypotheticals, while another 314 are associated with cell wall processes. A sequence analysis based prediction of trans-membrane helices , indicated that 850 proteins contained at-least one membraneassociated segment. Of these, structural models are available for 251, many of which participate in cell-wall processes. Further, of the membrane associated proteins, the cytochrome C Abmole Perifosine oxidase subunitI-like fold occurs most often , followed by glycerol-3-phosphate transporter and SNF like folds, . Among the least commonly occurring folds, 9 folds are associated with extracellular processes such as toxin-anti toxin system and that are implicated in cell adhesion. Functional annotation has been carried out by broadly three different approaches: first through known fold-function associations, second through identification of known sub-structural motifs and third through binding site identification, comparison with sites of known ligands and a subsequent ligand association. Fold to function assignments have been possible for 2832 different proteins, using the classification scheme reported earlier . Seven broad functional categories associated to 50 different functions have been linked to the models of TB. Validation for this part of the annotation pipeline has been carried out on the 312 protein test set. Binding site prediction, comparison and ligand association computed using PocketDepth and PocketMatch were compared with that in their corresponding crystal structures , which indicated that ligand associations and hence annotations were in most cases correct, when the detected similarity was atleast moderately high . Similarly a comparison between the fold-based function annotation and the GO terms, also indicated broad agreement . While the annotations and a broad functional PLX-4720 chemical information category were available in the databases for many proteins based on literature and sequence analyses, several new associations have been possible through modeling and structural analyses, some examples are described later. Of the 639 conserved proteins annotated as hypothetical proteins in the TB genome, 560 proteins can now be associated with fold-based function annotation through the pipeline reported here. While 282 of these 639 proteins have an association with one or more functional categories of COG, our protocol has provided an annotation for 560 proteins through structural analysis.