Computational protocol: Phyllostomid bat microbiome composition is associated to host phylogeny and feeding strategies

Similar protocols

Protocol publication

[…] We followed the protocol described by Caporaso et al. () for paired-end 16S rRNA gene community sequencing using primers 515F/806R that target the hypervariable region V4 in both bacteria and archaea. Intestine regions per individual and species were treated as a separate sample, and each PCR included a specific Golay reverse primer (Caporaso et al., ). DNA concentrations were calculated from each sample with a Qubit dsDNA assay (Invitrogen, Carlsbad, CA). On average 2 ng/μl of total DNA were added to each PCR reaction, of a total volume of 25 μl, and had 2.5 μl Takara (TaKaRa Corp., Shiga, Japan) ExTaq PCR 10X buffer, 2 μl Takara dNTP mix (2.5 mM), 0.7 μl bovine serum albumin (20 mg/ ml, Roche), forward and reverse primers (10 mM final concentration), 0.125 μl Takara Ex Taq DNA Polymerase (5u/μl) and nuclease free-water. The amplification protocol included an initial denaturalization step at 95°C for 3 min, followed by 35 cycles of 95°C, 30 s, 52°C, 40 s, 72°C, 90 s, and a final extension at 72°C for 12 min. Each sample was amplified in triplicate, combined and purified using the SPRI magnetic bead, Agencourt AMPure XP PCR purification system (Beckman Coulter, Brea, CA, USA). DNA concentration after pooling the PCR products for each sample and purification steps were obtained with the Qubit dsDNA HS assay. Amplicons were pooled (~20 ng per sample) and sequenced on Illumina MiSeq platform (at the Yale Center for Genome Analysis, CT, USA), resulting in ~250 bp paired end reads. The sequence data are available from BioProject ID: PRJNA260412.Paired-end sequences were overlapped and merged using FLASH (Magoč and Salzberg, ). Quality filtering and demultiplexing were performed in QIIME (r = 1; p = 0.75; q = 3; n = 0,–min_count 0.005%) as described previously (Caporaso et al., ; Bokulich et al., ). For L. yerbabuenae two samples did not pass the quality filtering and were excluded from further analysis. Sequences were grouped into operational taxonomic units (OTUs) based on 97% sequence identity, and chimeric sequences were removed, using USEARCH (Edgar, ). OTUs were given taxonomic assignments in QIIME (Caporaso et al., ) version 1.7.0 using RDP classifier (Wang et al., ) and Greengenes database release 13_5. Phylogenetic trees were created using FastTree2 (Price et al., ) under QIIME's default parameters and these trees were used for the calculation of α (Shannon's H', Fisher's and Faith's PD) and β diversity (weighted UniFrac distance) metrics. Communities were standardized to a total number of 12,000 sequences per intestine region per individual, or intestine regions were combined into 44,600 sequences per individual within each species. The weighted UniFrac distance matrices were used to visualize microbiome composition within bat species. [...] All statistical tests were conducted by using R packages “ade4” (Chessel et al., ) and “vegan” in the R statistical environment (Oksanen et al., ). Results are defined to be significant at P < 0.05. Correlation amongst bat species, intestinal regions and changes in microbial community abundances were explored via canonical correlation analysis as implemented in ade4. To test differences between α diversity of different bat species with different feeding strategies, we used One-Way ANOVA followed by the Tukey's honestly significant difference test. Ordination of the whole community detected by 16S rRNA gene amplicon sequencing was created from UniFrac matrix calculated by QIIME software and presented in a principal coordinates analysis plot. The contribution of feeding strategy, host species, age, sampling site and sex to β-diversity was tested via permutational MANOVA model as implemented in the “adonis” function of the vegan package in R. For this analysis first each parameter was sequentially added to the model. Secondly, group variations were controlled amongst feeding strategies and bat species. The evolutionary relationships among Phyllostomidae included in this study were inferred from mitochondrial CytB sequence identities. CytB sequences were aligned with Muscle (Edgar, ) and the calculated pairwise distances were used for clustering by UPGMA. […]

Pipeline specifications

Software tools QIIME, USEARCH, RDP Classifier, FastTree, UniFrac, vegan, MUSCLE
Databases Greengenes
Applications Phylogenetics, 16S rRNA-seq analysis, Nucleotide sequence alignment
Organisms Desmodus rotundus, Carollia perspicillata, Artibeus jamaicensis