The role and risks of selective adaptation in extreme coral habitats

This electronic notebook provides the scripts employed to analyze gene expression dynamics of the coral Porites lutea inhabiting mangrove and reef sites in Woody Isles, Australia.

Published in Nature Communications.

RNA-Seq reads quality filtering and mapping

Quality filtering and mapping - RNA-Seq reads processing included adapter trimming using Cutadapt v2.6 (Martin, 2011) and quality filtering using Trimmomatic v0.39 (Bolger et al., 2014). Reads were aligned to the P. lutea genome assembly using HISAT2 v2.2.1 (Kim et al., 2019). Transcripts assembly and quantification were performed using Stringtie v2.2.5 (Pertea et al., 2015).

Coral and algal symbiont species identification

Symbiont species identification - High quality reads were BLASTed using Diamond v2.0.11 (Buchfink et al., 2021) against the NCBI, Reefgenomics, Marinegenomics and UQ eSpace proteomes databases of Symbiodiniaceae and stony coral species.

SNPs characterization

Coral host SNPs characterization - Single nucleotide polymorphisms (SNPs) analysis was conducted using the Genome Analysis Toolkit framework (GATK, v4.2.0; (McKenna et al., 2010)) following the recommended RNA-Seq SNPs practice of the Broad Institute ((Auwera et al. 2013)), with necessary adjustments for genotype calling in non-model organisms where variants sites are not known beforehand. HISAT-aligned reads were sorted and marked for duplicates, variant calling was performed with the GATK HaplotypeCaller tool (McKenna et al., 2010) and genotypes were then jointly called using the GATK GenotypeGVCFs tool. The GATK SelectVariants and VariantFiltration tools were used to filter the joined variant-calling matrices for quality by depth. Filtering for linkage disequilibrium was carried out using PLINK (v2.0, (Purcell et al. 2007). To assess genetic differentiation among mangroves and reef corals, the dissimilarity coefficient and the proportions of shared ancestral alleles were estimated by computing pairwise identity-by-state distances using the SNPRelate R package v1.20.1,69 (Zheng et al., 2012). In addition, the fixation index (Fst)(Weir & Cockerham, 1984) was estimated using the R package HIERFSTAT v0.5.10.

Differential expression

Coral host differential expression - DE analysis was conducted using Bioconductor DEseq2 v1.26.0 (Love et al., 2014) in the R environment (v3.6.3) considering a single factor (location).

Functional annotation

Coral host functional annotation - Functional annotation of P. lutea protein sequences was performed using Diamond v2.0.11 (Buchfink et al., 2021), Uniprot and eggNOG.

Gene ontology enrichment

Coral host gene ontology enrichment analysis - GO enrichment analysis was performed using the package Goseq (v1.42.0; Young et al. 2010) in the R environment. For the resulting enriched GO terms, slim categories were obtained using the goSlim function of the R package GSEABase (v1.52.1) using the GOslim generic obo v1.2,65 as reference database .