🪸 Coral Multi-'Omics Explorer

Open the KEGG global metabolic-pathways map with this dataset's KEGG orthologs (KOs) overlaid — shaded by representation (darker = more genes mapped to that reaction). Rendered on KEGG; for the complete KO set, download the list and paste it into KEGG Mapper's "Search&Color Pathway".

No dataset loaded.

FASTA & TSV honor the filters set on the Expression Table tab (compartment, molecule, DE, and search) and the active comparison. The client export covers the contigs loaded in the browser (top by variance + all DE). For the complete filtered set (e.g. every host mRNA), use the server export when the backend is online: Server FASTA (full set)

Complete, gzipped nucleotide FASTA for an entire compartment of an experiment — e.g. every Symbiodiniaceae (endosymbiont) sequence from S. hystrix. Hosted directly; no backend needed.

Complete DIAMOND blastx of the P. acuta PDpCO₂ de novo transcriptome against NCBI nr — the top hit per contig with % identity, alignment length, e-value, bit-score, source organism, and description. 239,914 contigs annotated (of 385,298; the rest have no nr homolog). Per-contig hits are already shown in the Expression Table and gene popups; this is the full set. TSV, gzipped.

The host genotype is the PDLTTE mitochondrial ORF (mORF) alignment vs P. damicornis mitogenomes. qPCR primer/sequence files are also linked inline (🧫 qPCR) from the Expression Table for genes with a designed assay.

Complete predicted-protein FASTA for each reference genome/transcriptome (gzipped) — the full genome-derived sets, including proteins not detected by MS.

Nucleotide transcriptomes (mRNA):

Nucleotide genome assemblies are at their original sources: P. acuta — Vidal-Dupiol et al. 2019; S. fitti — Reich et al. 2021; A. cervicornis — Baums lab.

Original supplementary data files from the published studies.

Raw mass-spectrometry files are large and are archived in public proteomics repositories rather than re-hosted here:

Each JSON file should have this top-level structure. Per-contig fields:

{
  "species": "Seriatopora hystrix",
  "study": "SHVTS",
  "comparison": "Variable vs. Stable thermal regime",
  "total_contigs": 209481,
  "samples": ["HBH_V_1","HBH_V_2","HBH_V_3","HWN_S_1","HWN_S_2","HWN_S_3"],
  "sample_meta": [{"name":"HBH_V_1","site":"Houbihu","condition":"Variable","rep":1}, ...],
  "contigs": [
    {
      "id": "comp12345_c0_seq1",
      "lfc": 2.34,          // log2 fold change (Variable / Stable)
      "pval": 0.0012,        // p-value (Welch t-test or DESeq2 padj)
      "nlp": 2.92,           // -log10(pval) — for volcano y-axis
      "ms": 12.4,            // mean FPKM in Stable condition
      "mv": 55.1,            // mean FPKM in Variable condition
      "var": 180.2,          // variance across all samples
      "fpkm": [44,60,62, 10,14,15],  // per-sample FPKM (order matches "samples")
      "top_blast_hit": "heat shock protein 70 [Acropora digitifera]",
      "blast_evalue": 1.2e-45,
      "gene_name": "hsp70",
      "kegg_ids": "K04043",  // semicolon-separated if multiple
      "go_terms": "GO:0006950;GO:0051082",
      "annotation": "Stress response / chaperone"
    }
  ],
  // Optional: multiple DE contrasts (e.g. PDLTTE 2×2 factorial)
  "contrasts": [
    {
      "name": "High vs. Control — Week 2",
      "comparison": "30°C vs. 27°C at week 2",
      "contigs": [ ...same schema... ]
    },
    {
      "name": "High vs. Control — Week 36",
      "comparison": "30°C vs. 27°C at week 36",
      "contigs": [ ... ]
    }
  ]
}

Expression values are FPKM (Fragments Per Kilobase of transcript per Million reads mapped), quantified with RSEM. Differential expression in this explorer uses Welch two-sample t-tests between user-selected conditions; log₂ fold change uses a pseudocount of 0.1 to handle zero-FPKM values. KEGG annotations via KOBAS; functional annotations via Pfam-HMM and HMMER. Protein sequences from TransDecoder ORF prediction (6-frame translation fallback when .pep unavailable). BLAST search (blastn, blastp, blastx, tblastn, and tblastx) requires the local server to be running with pre-built databases.

This explorer was built by Coral Reef Diagnostics. The original interactive websites for these datasets (hosted at the Institute of Information Science, Academia Sinica, Taiwan) are no longer online. This site aims to restore full queryability of both transcriptomes. Source data are from the published supplementary materials and original sequencing files.

• Differential expression: main effects by Welch two-sample t-test on log₂(FPKM+0.1); the temperature×time / site×treatment / temperature×CO₂ interaction term by two-way ANOVA (balanced designs) or an OLS interaction t-test (unbalanced PDpCO₂). "DEG" = |log₂FC|≥1 and p<0.05.
• Compartment classification (host / Symbiodiniaceae / microbe): every contig is binned by sequence homology — eggNOG / NCBI-nr best-hit lineage (host = Metazoa; symbiont = Alveolata, Stramenopiles, or green plants, sensu Mayfield et al. 2014; microbe = Bacteria/Archaea/Virus) supplemented by DIAMOND blastx against species-matched host + Symbiodiniaceae proteomes for contigs lacking an annotation. Contigs with no homology default to host.
• Host : Symbiodiniaceae normalization (toggle): each sample's host-gene values are divided by its host:Symbiodiniaceae contig ratio, and Symbiodiniaceae-gene values by the inverse (Symbiodiniaceae:host), then rescaled to the cross-sample mean — so compartment-heavy samples are down-weighted, controlling for differing symbiont load (sensu Mayfield et al. 2014).
• GO / KEGG enrichment: hypergeometric over-representation of the DEG set against the annotated dataset background, Benjamini–Hochberg FDR.
• Sample ordination (PCA): principal components of the top-variance genes (log₂ FPKM, gene-centered).
• Cross-experiment orthologs: genes linked across experiments by shared KEGG ortholog (KO) and eggNOG gene name.
• Proteomics: iTRAQ abundances normalized to the pooled reference channel; qPCR normalized to Solaris & genome-copy-proportion spikes; on-the-fly peptide search via MS-GF+ (1% FDR).

• Hypothetical proteome — every protein the organism is predicted to make: the full conceptually-translated reference (e.g. O. faveolata host 73,093; in-hospite Symbiodiniaceae 76,145). For the corals/symbionts assembled de novo, these are ORFs translated from the reference transcriptome; for A. cervicornis / S. fitti they come from the Baums reference genomes.
• Sequenced — proteins actually identified by mass spectrometry (master protein at High/Medium FDR), pooled across all iTRAQ/TMT batches of a study. This is the headline count on each proteome panel.
• iTRAQ-quantified only (toggle) — the far smaller subset of sequenced proteins that additionally carry quantitative iTRAQ reporter-ion ratios. Ticking the box filters the table and headline count to these.
• Composite Symbiodiniaceae reference: the in-hospite Symbiodiniaceae proteome is a composite, conceptually translated from multiple co-assembled transcriptomes — the symbiont community spans more than one genus (e.g. Breviolum + Durusdinium in O. faveolata), so transcripts from several Symbiodiniaceae are represented together.

Please cite the original study for any dataset you use (DOIs above), and you may additionally acknowledge this explorer:

Coral Multi-'Omics Explorer, Coral Reef Diagnostics (coralreefdiagnostics.com), accessed [date].

Data availability: PDLTTE assembly & expression — NCBI BioProject PRJNA227785 and Dryad 10.5061/dryad.rh04m; SHVTS — Mol. Ecol. 2016 supplements; SHVTS iTRAQ proteome — MassIVE MSV000085863 (cross-listed at NOAA NCEI); PDpCO₂ adult-coral iTRAQ proteome — MassIVE MSV000085868; P. acuta reference genome — Vidal-Dupiol et al. 2019 (bioRxiv 698688); S. fitti genome — Reich et al. 2021 (Mol. Ecol.)