Detection of Parasites in Microbiomes using Metagenomics

Despite a yearly death toll of up to one million people due to parasite-related infections, parasites are still neglected in genomics research. While there is progress in the detection of bacteria and viruses using metagenomics in the context of infectious diseases, there are still challenges in metagenomics-based detection of parasites. Here, we implement a workflow for the detection of parasites from metagenomics data. We employ stringent cut off criteria to limit false positive detections. We analysed a total of 7.120 metagenomics samples of which 359 originated from gut microbiomes of livestock (pigs and chicken) from nine countries, and 6.761 from gut microbiomes of humans (adults and infants) from 25 countries. Five parasite-related genera were detected in livestock, of which Blastocystis sp. was detected in 71% of all pig herds and Eimeria in 83% of all chicken flocks. Distinct gut bacterial taxa were associated with Blastocystis sp. abundance in pigs. Nine parasite-related genera were detected in humans. Blastocystis sp. subtypes ST1, ST2, and ST3 were detected in all countries, and ST3 was most predominant. A higher overall prevalence of Blastocystis sp. was observed in low-income countries as compared to high-income countries, and a higher diversity of Blastocystis subtypes (ST1, ST2, ST3, ST4, ST6, ST7, ST8) was detected in high-income countries as compared to low-income countries. The prevalence of Blastocystis sp. in infant gut microbiome samples was lower as compared to adults. Overall, metagenomics-based analysis may be a promising tool for parasite detection from complex microbiome samples in clinical and veterinary medicine. Metagenomics could become the preferred method for parasite detection for a wide range of biological samples. Current parasite detection methods often rely on microscopic examination of the sample or using specific PCR. Metagenomics-based analyses may allow for a faster and more convenient way of detecting parasites in humans and animals, as this approach could serve as a one-for-all untargeted approach for pathogen detection, including bacteria, viruses, and parasites.