Disruption of healthy microbial communities has been linked to numerous diseases, yet microbial interactions are little understood. This is due in part to the large number of bacteria, and the much larger number of interactions (easily in the millions), making experimental investigation very difficult at best and necessitating the nascent field of computational exploration through microbial correlation networks. We benchmark the performance of eight correlation techniques on simulated and real data in response to challenges specific to microbiome studies: fractional sampling of ribosomal RNA sequences, uneven sampling depths, rare microbes and a high proportion of zero counts. Also tested is the ability to distinguish signals from noise, and detect a range of ecological and time-series relationships. Finally, we provide specific recommendations for correlation technique usage. Although some methods perform better than others, there is still considerable need for improvement in current techniques.

BackgroundIdentifying viral sequences in mixed metagenomes containing both viral and host contigs is a critical first step in analyzing the viral component of samples. Current tools for distinguishing prokaryotic virus and host contigs primarily use gene-based similarity approaches. Such approaches can significantly limit results especially for short contigs that have few predicted proteins or lack proteins with similarity to previously known viruses.MethodsWe have developed VirFinder, the first k-mer frequency based, machine learning method for virus contig identification that entirely avoids gene-based similarity searches. VirFinder instead identifies viral sequences based on our empirical observation that viruses and hosts have discernibly different k-mer signatures. VirFinder’s performance in correctly identifying viral sequences was tested by training its machine learning model on sequences from host and viral genomes sequenced before 1 January 2014 and evaluating on sequences obtained after 1 January 2014.ResultsVirFinder had significantly better rates of identifying true viral contigs (true positive rates (TPRs)) than VirSorter, the current state-of-the-art gene-based virus classification tool, when evaluated with either contigs subsampled from complete genomes or assembled from a simulated human gut metagenome. For example, for contigs subsampled from complete genomes, VirFinder had 78-, 2.4-, and 1.8-fold higher TPRs than VirSorter for 1, 3, and 5 kb contigs, respectively, at the same false positive rates as VirSorter (0, 0.003, and 0.006, respectively), thus VirFinder works considerably better for small contigs than VirSorter. VirFinder furthermore identified several recently sequenced virus genomes (after 1 January 2014) that VirSorter did not and that have no nucleotide similarity to previously sequenced viruses, demonstrating VirFinder’s potential advantage in identifying novel viral sequences. Application of VirFinder to a set of human gut metagenomes from healthy and liver cirrhosis patients reveals higher viral diversity in healthy individuals than cirrhosis patients. We also identified contig bins containing crAssphage-like contigs with higher abundance in healthy patients and a putative Veillonella genus prophage associated with cirrhosis patients.ConclusionsThis innovative k-mer based tool complements gene-based approaches and will significantly improve prokaryotic viral sequence identification, especially for metagenomic-based studies of viral ecology.Electronic supplementary materialThe online version of this article (doi:10.1186/s40168-017-0283-5) contains supplementary material, which is available to authorized users.

Microbes have central roles in ocean food webs and global biogeochemical processes, yet specific ecological relationships among these taxa are largely unknown. This is in part due to the dilute, microscopic nature of the planktonic microbial community, which prevents direct observation of their interactions. Here, we use a holistic (that is, microbial system-wide) approach to investigate time-dependent variations among taxa from all three domains of life in a marine microbial community. We investigated the community composition of bacteria, archaea and protists through cultivation-independent methods, along with total bacterial and viral abundance, and physicochemical observations. Samples and observations were collected monthly over 3 years at a welldescribed ocean time-series site of southern California. To find associations among these organisms, we calculated time-dependent rank correlations (that is, local similarity correlations) among relative abundances of bacteria, archaea, protists, total abundance of bacteria and viruses and physico-chemical parameters. We used a network generated from these statistical correlations to visualize and identify time-dependent associations among ecologically important taxa, for example, the SAR11 cluster, stramenopiles, alveolates, cyanobacteria and ammonia-oxidizing archaea. Negative correlations, perhaps suggesting competition or predation, were also common. The analysis revealed a progression of microbial communities through time, and also a group of unknown eukaryotes that were highly correlated with dinoflagellates, indicating possible symbioses or parasitism. Possible 'keystone' species were evident. The network has statistical features similar to previously described ecological networks, and in network parlance has non-random, small world properties (that is, highly interconnected nodes). This approach provides new insights into the natural history of microbes.

The recent development of metagenomic sequencing makes it possible to sequence microbial genomes including viruses in an environmental sample. Identifying viral sequences from metagenomic data is critical for downstream virus analyses. The existing reference-based and gene homology-based methods are not efficient in identifying unknown viruses or short viral sequences. Here we have developed a reference-free and alignment-free machine learning method, DeepVirFinder, for predicting viral sequences in metagenomic data using deep learning techniques. DeepVirFinder was trained based on a large number of viral sequences discovered before May 2015. Evaluated on the sequences after that date, DeepVirFinder outperformed the state-of-the-art method VirFinder at all contig lengths. Enlarging the training data by adding millions of purified viral sequences from environmental metavirome samples significantly improves the accuracy for predicting underrepresented viruses. Applying DeepVirFinder to real human gut metagenomic samples from patients with colorectal carcinoma (CRC) identified 51,138 viral sequences belonging to 175 bins. Ten bins were associated with the cancer status, indicating their potential use for non-invasive diagnosis of CRC. In summary, DeepVirFinder greatly improved the precision and recall rates of viral identification, and it will significantly accelerate the discovery rate of viruses.