Two step pcr protocol

The following procedure was done according to Tallei et al. [16 (link)]. An Illumina two-step PCR protocol was used for preparing the amplicons library. The first stage was to generate PCR products of V3-V4 regions using Nextera-style tag sequences (overhang sequences) with the following sequences: forward overhang P5-tag: 5′TCGTCGGCAGCGTCAGATGTGTAT AAGAGACAG-[locus-specific sequence] and reverse overhang P7-tag: 5′GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG-[locus-specific sequence]. The second stage used sample-specific Illumina Nextera XT dual indices (Nextera XT i7 index and Nextera XT i5 index) with the following sequences: P5-PCR index primer: 5′AATGATACGGCGACCACCGAGATCTACAC[i5]TCGTCGGCAGCGTC and P7-PCR index primer:5′CAAGCAGAAGACGGCATACGAGAT[i7]GTCTCGTGGGC TCGG. The final products were assessed using TapeStation 4200 from Agilent Technologies, HelixyteTM green dsDNA quantifying reagent, and qPCR using Jetseq library quantification Lo-Rox kit from Bioline (London, UK). The paired-end sequences were generated in a 2 × 300 bp format from MiSeq.

Bacterial metabarcoding was performed using 16S rRNA gene amplicon sequencing. Libraries were generated using the Illumina two-step PCR protocol targeting the V3-V4 region [26 (link)]. A total of 252 libraries (six families × seven sampling time points × three replicates × two infectious environments) were paired-end sequenced with a 2 × 250 bp read length at the Genome Quebec platform (RRID: SCR_017703) on a MiSeq system (Illumina) according to the manufacturer’s protocol. A total of 41,012,155 pairs of sequences were obtained. Metabarcoding data was processed using the FROGS pipeline [28 (link)]. Briefly, paired reads were merged using FLASH [29 (link)]. After cleaning steps and singleton filtering, 26,442,455 sequences were retained for further analyses. After denoising and primer/adapter removal with CUTADAPT, clustering was performed using SWARM, which uses a two-step clustering algorithm with a threshold corresponding to the maximum number of differences between two Operational Taxonomic Units (OTU) (denoising step d = 1; aggregation distance = 3) [30 ]. Chimeras were removed using VSEARCH [31 (link)]. Resulting OTUs were affiliated using Blast + against the Silva database (release 128).