Assessment of sequence descriptions of selected Theileria parva hypothetical proteins retrieved from sequence similarity search databases

Abstract

The protozoan parasite Theileria parva is the causative agent of cattle theileriosis, a disease with a destructive impact on the agricultural economy through mortality and morbidity of affected cattle. In cattle, T. parva infection results in varied disease syndromes depending on the parasite host of origin; cattle-derived T. parva causes East Coast fever while buffaloderived parasites cause Corridor disease. The differences in the resulting disease caused by T. parva infection have raised an interest to understand the proteins involved in the disease manifestations. Consequently, a transcriptome study comparing the cattle and buffalo-derived T. parva isolates was undertaken; differentially expressed genes were detected of which 74% (867) were hypothetical proteins (HPs). Since HPs could play a vital role in the pathogenicity and host-parasite interaction, the primary aim of the study was to identify biological roles of these proteins. A combination of in silico analysis tools was employed to annotate HPs according to sequence descriptions, confirmed by sequence homology in comparison with closely related species and conserved domains. Initial screening for sequence descriptions (SDs) based on sequence similarity search using Blast2GO retrieved results for 392 HPs. Comparison of this output to other databases (KEGG and KOBAS) detected consensus SDs for 229 HPs, of which 109 were further confirmed by inferring homology to related species. Sequence homology analysis also resulted in designation of SDs to 74 HPs from the remaining 163 without consensus SDs from database analyses. For HPs which did not meet the criteria employed in sequence homology analysis (209), conserved domain analysis facilitated assigning of SDs for 114 HPs. Overall, 297 (76%) HPs were successfully allocated SDs. Finally, the results from this study have showed that output from automated sequence similarity databases is not always reliable in assigning SDs for specific species, making confirmation using other approaches necessary.