Non-random patterns in viral diversity

It is currently unclear whether changes in viral communities will ever be predictable. Here we investigate whether viral communities in wildlife are inherently structured (inferring predictability) by looking at whether communities are assembled through deterministic (often predictable) or stochastic (not predictable) processes. We sample macaque faeces across nine sites in Bangladesh and use consensus PCR and sequencing to discover 184 viruses from 14 viral families. We then use network modelling and statistical null-hypothesis testing to show the presence of non-random deterministic patterns at different scales, between sites and within individuals. We show that the effects of determinism are not absolute however, as stochastic patterns are also observed. In showing that determinism is an important process in viral community assembly we conclude that it should be possible to forecast changes to some portion of a viral community, however there will always be some portion for which prediction will be unlikely.

correlate results for each. General agreement is observed and Jaccard index is selected for incorporation in the null model (see Fig 3).

Supplementary Figure 2:
Histogram showing the distribution of β-Nearest Taxon Index (β-NTI) values. If the β-NTI distribution is significantly shifted towards -2 or +2, it would indicate an important role for deterministic selective forces. Here we show that the distribution is not significantly shifted from zero, which in combination with the Jaccard-based null modelling (see Fig 3) suggests that non-random patterns may be emerging due to dispersal limitation.
Supplementary Figure 3: Schematic of Universal Control 2. Two constructs (UC1 and UC2) were generated to provide positive control material for the consensus PCR screening. These constructs are made up of sequential (non-overlapping) primer-binding sites for all assays, interspersed with short stretches of synthetic sequence. UC2 is shown here as an example to illustrate the structure of these controls. UC1 and UC2 allow the amplification of PCR products of the expected size in order to confirm successful execution of the assay. They also allow easy recognition of sample contamination, given that amplified products contain a series of primer-binding sites rather than a real viral sequence. Protocol P-031 (herpesvirus) is shown as an example.
Constructs include a T7-promoter sequence at the beginning for transcription into RNA templates where desired. Description of all viruses discovered in macaques: A total of 184 viruses were identified using a combination of consensus PCR (cPCR) and high-throughput sequencing (HTS). All viruses identified by HTS (contigs or singletons) were subsequently confirmed by PCR or cPCR and the length of this confirmed sequence is indicated. For some viruses, putative partial or near complete genomes were assembled from the HTS data (*). Note: a large diversity of picobirnaviruses (PbVs) was discovered in these macaques. Initially, approximately 50 different PbV sequences were detected in the HTS data. An alignment of these sequences was then used to design a novel cPCR assay (see P-030 in Supplemental Table 4), which was used to re-screen all samples individually by cPCR for the detection of all 120 viruses presented here. We therefore attribute the discovery of these viruses to HTS because no PbV assay used in the initial cPCR screen and none of them would have been discovered without the initial HTS data -but also note that the full diversity of PbVs discovered is actually the result of a combined HTS/cPCR strategy.

Supplemental Methods: Consensus PCR (cPCR) Assays
Below is a detailed description of all cPCR assays used in this study, accompanied by anecdotal notes that may be useful to investigators wishing to use these methods for their own purposes.

NOTES:
This assay works well to detect novel astroviruses, however investigators should be aware that it does cross-react fairly readily and often produces false-positive bands of the expected size. Sequencing will be required to rule out false positives.  Lozano 3 l, which describes a method for full sequence characterization of arenaviruses. Two sets of primers have been selected from the paper (P-003 here, and P-004 below).  Lozano 3 ,, which describes a method for full sequence characterization of arenaviruses. Two sets of primers have been selected from the paper (P-003 above, and P-004 here). A slightly modified version of the 7C primer has been designed based on a more recent alignment of arenaviruses. NOTES: Primer sequences have been modified to increase the ability of the assay to detect widely variant CoVs. A second, hemi-nested step has also been added to increase sensitivity. Nested PCR should be performed using both forward primers.

NOTE:
Designed as a one-step RT-PCR. Using synthetic standards, the sensitivity was calculated to be 50-500 copies, so even with one round of PCR, it should be quite sensitive. The primers are well validated for known filoviruses, but their lack of degeneracy might prevent detection of diverse or novel filoviruses.

NOTE:
Originally developed as a real-time (SYBR green) one-step RT-PCR method for the universal detection and identification of flaviviruses. However works well in a conventional PCR platform. The assay can be easily inhibited by too much nucleic acid, giving false negative results. Be sure to use <500ng/ul cDNA at all times.

NOTE:
This assay has not been as extensively validated as P-009, especially on clinical samples. However, it does contain a nested primer pair. In theory this should improve sensitivity, however the large size of the first round product may reduce the overall efficiency of this PCR. The size of the second round amplicon is very small. While this is still sufficient to confirm a positive result, we encourage investigators to run the PCR reactions from both rounds (on gels) as far more information can be gathered from the larger (round 1) product, if successfully amplified.

NOTE:
The paper provides good validation of the primers against various paramyxoviruses, but when applied to clinical samples this assay can cross-react with host sequences quite readily. It has been observed that this is a particular problem with bat samples. When a paramyxovirus is present in a sample, the assay does seem to work well, and (mostly) amplifies a single and specific product, with very little non-specific amplification. Investigators should therefore proceed cautiously with any sample that looks to have products that are close to the right size, but which also have a high background of non-specific amplification. It is probable that such samples are in fact negative. NOTE: This paper presents multiple assays, both real-time and conventional. Here we present a conventional assay that should be broadly reactive for all henipaviruses, and which is hemi-nested for increased sensitivity. While focused on henipaviruses, investigators should note that this assay is not specific for henipaviruses, and will also pick up other paramyxoviruses.

NOTES:
The first primer pair will often work well enough on their own, but there is a nested set included to increase sensitivity where required (recommended). Please note that these primers can, on occasion, produce a non-specific product (host) of the expected size. All positives must be confirmed by sequencing. NOTES: This paper contains primer sets for all three segments of the bunyaviruses. It is recommended that investigators use at least two of these when screening for orthobunyaviruses (see also S above).

REFERENCE: Reid, S.M. et al. (1999) 20 TARGET:
Capsid-coding region CONTROLS: Calicivirus cDNA ENZYME: QIAGEN NOTES: This assay was designed to be more degenerate than P-025 above. It was developed to capture the highly diverse bat influenza viruses (recently discovered) in addition to all 'classic' influenza A viruses. While a nested assay is presented here, we have to date not observed any increased sensitivity from the nested round. NOTES: Very well tested assay, and shown to work well on clinical samples. This assay has a bias for gammaherpesviruses. It will still detect alpha and beta herpesviruses, but if there is a co-infection between a gamma and either alpha or beta, it will preferentially amplify the gamma (so investigators should always consider cloning to identify co-infections). Important: this assay requires the addition of dimethyl sulphoxide (DMSO) to final dilution of 5% total volume (e.g. 1l in a 20l reaction). NOTES: This is a paper describing several poxvirus PCRs. The protocol described here is the polymerase assay. The topoisomerase assay is described below (P-035). Important: this assay requires the addition of dimethyl sulphoxide (DMSO) to final dilution of 5% total volume (e.g. 1l in a 20l reaction). NOTES: This is a paper describing several poxvirus PCRs. The protocol described here is the topoisomerase assay. The polymerase assay is described above (P-034). Important: this assay requires the addition of dimethyl sulphoxide (DMSO) to final dilution of 5% total volume (e.g. 1l in a 20l reaction). . NOTES: This assay has been widely used to detect novel adenoviruses in animals, including many bat, bird and reptilian adenoviruses, from all four genera. They have even been used to identify a putative new genus in the family. Please note: we have tested this protocol with multiple enzymes, and only seems to work reliably with AmpliTaq Gold.

NOTES:
An assay originally used to identify novel polyoma in chimps. But was designed to be broadly reactive, and we have used it successfully to identify several novel polyomaviruses in diverse species (including dolphins).