Detecting Bacterial Pathogens and Antibiotic Resistance Genes in Wastewater Treatment Plants Using Annotations and Bioinformatics

Abstract

Limited water resources and high consumption rates present the United Arab Emirates with a significant sustainability challenge. In recent years, treated wastewater is reused as an alternative water resource in Abu Dhabi Emirate both for landscaping and recreation areas. The wider use of treated wastewater will contribute to the conservation of conventional water resources, but raises concerns regarding the safeguarding of both human health and the environment itself. The aim of this study is to develop a pipeline to detect bacterial pathogens and Antibiotic Resistance Genes (ARGs) present in treated wastewater samples. Bacterial pathogens and ARGs detection was carried out after predicting the full metagenome from the 16S rRNA Illumina sequencing via PICRUSt as 16S rRNA profile is not capable to provide microbial metabolism and functionality. A pipeline was built to detect pathogens and ARGs resulted in detecting 93 pathogenicity genes and five ARGs. Various human, mammal, and plant pathogenic genes were present in the collected treated wastewater samples. Detected ARGs abundances in the predicted full metagenomes of treated wastewater samples were compared to ARGs abundances in environmental samples, and it was noticed that most ARGs in treated wastewater samples have relatively lower abundances compared to the environmental samples, while few ARGs have slightly higher abundances compared to the environmental samples. Furthermore, the accuracy of full metagenome prediction was investigated via obtaining fully sequenced metagenome from MG-RAST server and extracting 16S rRNA profile. Hence, a predicted metagenome using extracted 16S rRNA was constructed via PICRUSt. Pathogens and ARGs detection pipeline for 16S rRNA and metagenomes alike were built using ontological annotations and combinations of databases. It was noted that pathogens and ARGs detection in the predicted metagenome was inordinately high compared to the real full metagenome. The overprediction by PICRUSt can still be used as a screening method, if at least the number of False Negatives is low. Thus, results suggest the use of the full metagenome in order to study and detect microbial functionality and metabolism accurately. Moreover, further validation of the approach should be carried out and quantitative Polymerase Chain Reaction (qPCR) should be conducted to determine the exact concentration and genes abundances.

Date of Award	May 2015
Original language	American English
Supervisor	Andreas Henschel (Supervisor)