RT Journal Article
SR Electronic
T1 Evaluating the accuracy of DNA stable isotope probing
JF bioRxiv
FD Cold Spring Harbor Laboratory
SP 138719
DO 10.1101/138719
A1 Nicholas D. Youngblut
A1 Daniel H. Buckley
YR 2017
UL http://biorxiv.org/content/early/2017/05/17/138719.abstract
AB Originality-Significance Statement By combining DNA Stable Isotope Probing (DNA-SIP) with multiplexed high throughput DNA sequencing (HTS-DNA-SIP), it is now possible to identify patterns of isotope incorporation for thousands of microbial taxa. HTS-DNA-SIP has enormous potential to reveal patterns of carbon and nitrogen exchange within microbial food webs. A current limitation is that, due to the expense of these experiments, it has been impossible to evaluate the accuracy of DNA-SIP methods. We have developed a model that simulates DNA-SIP data, and we use the model to systematically evaluate and validate the accuracy of DNA-SIP analyses. This model can determine the analytical accuracy of DNA-SIP experiments in a range of contexts. Furthermore, the ability to predict experimental outcomes, as a function of experimental design and community characteristics, should be of great use in the design and interpretation DNA-SIP experiments.Summary DNA Stable isotope probing (DNA-SIP) is a powerful method that identifies in situ isotope assimilation by microbial taxa. Combining DNA-SIP with multiplexed high throughput DNA sequencing (HTS-DNA-SIP) creates the potential to map in situ assimilation dynamics for thousands of microbial taxonomic units. However, the accuracy of methods for analyzing DNA-SIP data has never been evaluated. We have developed a toolset (SIPSim) for simulating HTS-DNA-SIP datasets and evaluating the accuracy of methods for analyzing HTS-DNA-SIP data. We evaluated two different approaches to analyzing HTS-DNA-SIP data: “high resolution stable isotope probing” (HR-SIP) and “quantitative stable isotope probing” (q-SIP). HR-SIP was highly specific and moderately sensitive, with very few false positives but potential for false negatives. In contrast, q-SIP had fewer false negatives but many false positives. We also found HR-SIP more robust than q-SIP with respect to experimental variance. Furthermore, we found that the detection sensitivity of HTS-DNA-SIP can be increased without compromising specificity by evaluating evidence of isotope incorporation over multiple windows of buoyant density (MW-HR-SIP). SIPSim provides a platform for determining the accuracy of HTS-DNA-SIP methods across a range of experimental parameters, which will be useful in the design, analysis, and validation of DNA-SIP experiments.