Barcodes

DNA Extraction was carried out using the DNeasy® PowerSoil® Pro Kit (Qiagen) with 250 mg of homogenized soil from each sample. This study used a DNA amplification approach, as described by Supplementary methods in Caporaso et al. (2012) (link). The V4 region of the 16S rRNA gene was amplified together with barcodes (Illumina, San Diego, CA, United States) adaptor and sequencing primers, and the target primers, 515F and 806R. The efficiency of the amplification process was assessed through agarose gel electrophoresis. The PCR products were then pooled for each amplicon based on gel electrophoresis band strength and combined in equimolar concentrations.

Sequencing reads were demultiplexed first by Illumina barcodes, and then by custom ‘inline index’ barcodes unique to each sample replicate. Barcodes were trimmed using the fastx barcode splitter and trimmer tools [106 (link)]. The transposon was identified in a two-step pattern matching process, allowing for 3 and 1 nucleotide mismatch respectively. Reads that did not contain the transposon sequence were discarded, while reads with successful transposon-matching were trimmed of the transposon sequence and mapped to the reference genome (CP025209.1) using bwa mem [107 (link)]. The reference genome was closed and annotated by microbesNG, using Prokka (v 1.12; [108 (link)]). The plasmid deposited with this reference sequence was found to be an artefact and subsequently discarded from our analysis. Transposon insertion sequencing reads that successfully mapped to the reference genome were sorted and manipulated using samtools and bedtools [109 (link)]. The first nucleotide at the 52 end of each read was counted as the transposon insertion site. Data were viewed using the Artemis genome browser [110 (link)]. FASTQ data are available at the European Nucleotide Archive for download (accession: PRJEB56349). The processed insertion data can be viewed online at: https://tradis-vault.qfab.org/

We used Agencourt DNAdvance Genomic DNA Isolation Kit (Beckman Coulter, Inc.) to extract DNA from 163 backcross nymphs and their parents. A total of four offspring failed to generate enough reads for successful analyses and were eliminated.
We sequenced individuals in three Illumina Miseq lanes (Cornell Life Sciences Core Laboratory Center for first batch of individuals, Genewiz for second and third batch). For each individual, we first performed 5 PCR reactions with 24 primer pairs in each using the Qiagen Multiplex PCR Kit (Qiagen) to amplify a total of 120 SNP markers, the 165 individuals were divided into two Illumina lanes. To each individual we added Illumina barcodes (N501-520/S701-729) via PCR. All barcoded individuals from each batch (lane) were combined in a single mix and cleaned up using Agencourt Ampure XL beads (Beckman Coulter, Inc.). The third Illumina lane was used to get data for loci that failed to generate enough reads in the previous runs (23 loci).

Sequencing of the samples in our study was completed in collaboration with other researchers as part of the EMP (http://www.earthmicrobiome.org/). Our collaborators at EMP amplified and sequenced the V4 region of the 16s rRNA gene using the bacterial/archaeal primer pair 515F/860R and following previously published methods (Caporaso et al., 2012 (link)). Amplicons were fused to Illumina barcodes and sequencing was completed on an Illumina platform.