iNAP: An integrated network analysis pipeline for microbiome studies

Integrated network analysis pipeline (iNAP) is an online analysis pipeline for generating and analyzing comprehensive ecological networks in microbiome studies. It is implemented in two sections, that is, network construction and network analysis, and integrates many open-access tools. Network construction contains multiple feasible alternatives, including correlation-based approaches (Pearson's correlation and Spearman's rank correlation along with random matrix theory, and sparse correlations for compositional data) and conditional dependence-based methods (extended local similarity analysis and sparse inverse covariance estimation for ecological association inference), while network analysis provides topological structures at different levels and the potential effects of environmental factors on network structures. Considering the full workflow, from microbiome data set to network result, iNAP contains the molecular ecological network analysis pipeline and interdomain ecological network analysis pipeline (IDENAP), which correspond to the intradomain and interdomain associations of microbial species at multiple taxonomic levels. Here, we describe the detailed workflow by taking IDENAP as an example and show the comprehensive steps to assist researchers to conduct the relevant analyses using their own data sets. Afterwards, some auxiliary tools facilitating the pipeline are introduced to effectively aid in the switch from local analysis to online operations. Therefore, iNAP, as an easy-to-use platform that provides multiple network-associated tools and approaches, can enable researchers to better understand the organization of microbial communities. iNAP is available at http://mem.rcees.ac.cn:8081 with free registration.