================================================================================ BARNACLE TENSOR DECOMPOSITION: MULTI-SPECIES CORAL GENE EXPRESSION NARRATIVE ================================================================================ EXECUTIVE SUMMARY ----------------- This analysis reveals conserved and divergent gene expression patterns across three coral species (Acropora pulchra, Porites evermanni, and Pocillopora tuahiniensis) over four experimental time points using tensor decomposition. The barnacle algorithm identified 35 expression components, with the top 5 components by weight being: C27, C24, C13, C29, C25. TEMPORAL DYNAMICS ----------------- Gene expression across the coral species shows distinct temporal phases: • Early Peak (9 components): Components C1, C3, C6, C7, C12 • Stable (9 components): Components C2, C4, C16, C20, C23 • Late Peak (9 components): Components C8, C10, C11, C13, C18 • Late Mid Peak (5 components): Components C5, C9, C15, C22, C26 • Early Mid Peak (2 components): Components C14, C33 • Dynamic (1 components): Components C32 SPECIES-SPECIFIC PATTERNS ------------------------- The three coral species show both shared and unique expression signatures: • Acropora pulchra: Shows highest expression in 10 components • Porites evermanni: Shows highest expression in 14 components • Pocillopora tuahiniensis: Shows highest expression in 11 components KEY BIOLOGICAL PROCESSES ------------------------ Analysis of GO term enrichment reveals major biological themes: • regulation of transcription by RNA polymerase II [GO:0006357] • positive regulation of transcription by RNA polymerase II [GO:0045944] • homophilic cell adhesion via plasma membrane adhesion molecules [GO:0007156] • cytoplasmic translation [GO:0002181] • translation [GO:0006412] • positive regulation of gene expression [GO:0010628] • negative regulation of transcription by RNA polymerase II [GO:0000122] • positive regulation of DNA-templated transcription [GO:0045893] • cellular response to xenobiotic stimulus [GO:0071466] • negative regulation of neuron apoptotic process [GO:0043524] COMPONENT INTERPRETATIONS ------------------------- Component 27 (Weight: 1994.0, Pattern: stable) ================================================== Top GO Processes: positive regulation of transcription by RNA polymerase II, regulation of transcription by RNA polymerase II, homophilic cell adhesion via plasma membrane adhesion molecules Key Proteins: Early growth response protein 1 (EGR-1) (Zinc finger protein Krox-24), Myosin light chain kinase, smooth muscle (MLCK) (smMLCK) (EC 2.7.11.18) (Kinase-related protein) (KRP) (Telokin) [Cleaved into: Myosin light chain kinase, smooth muscle, deglutamylated form], Ecdysone-induced protein 74EF isoform A (ETS-related protein E74A) Component 24 (Weight: 1501.5, Pattern: stable) ================================================== Top GO Processes: negative regulation of neuron apoptotic process, signal transduction, cytoskeleton organization Key Proteins: Adenylyl cyclase-associated protein 2 (CAP 2), Wilms tumor protein 1-interacting protein homolog (WT1-interacting protein homolog), Myosin regulatory light polypeptide 9 (Myosin regulatory light chain 2, smooth muscle isoform) (Myosin regulatory light chain 9) Component 13 (Weight: 1435.0, Pattern: late peak) ================================================== Top GO Processes: homophilic cell adhesion via plasma membrane adhesion molecules, cytoplasmic translation, translation Key Proteins: Lactotransferrin (Lactoferrin) (EC 3.4.21.-) (Growth-inhibiting protein 12) (Talalactoferrin) [Cleaved into: Lactoferricin-H (Lfcin-H); Kaliocin-1; Lactoferroxin-A; Lactoferroxin-B; Lactoferroxin-C], Microtubule-actin cross-linking factor 1, isoforms 1/2/3/4 (Actin cross-linking family 7), Apolipoprotein B-100 (Apo B-100) [Cleaved into: Apolipoprotein B-48 (Apo B-48)] Component 29 (Weight: 1299.2, Pattern: stable) ================================================== Top GO Processes: positive regulation of transcription by RNA polymerase II, regulation of transcription by RNA polymerase II, intracellular signal transduction Key Proteins: Heat shock factor protein 1 (HSF 1) (Heat shock transcription factor 1) (HSTF 1), Kruppel-like factor 1 (Erythroid krueppel-like transcription factor) (Kruppel-like factor d), Heat shock protein HSP 90-beta Component 25 (Weight: 1281.5, Pattern: early peak) ================================================== Top GO Processes: cytoplasmic translation, translation, regulation of transcription by RNA polymerase II Key Proteins: Early growth response protein 1 (EGR-1) (Zinc finger protein Krox-24), Lactotransferrin (Lactoferrin) (EC 3.4.21.-) (Growth-inhibiting protein 12) (Talalactoferrin) [Cleaved into: Lactoferricin-H (Lfcin-H); Kaliocin-1; Lactoferroxin-A; Lactoferroxin-B; Lactoferroxin-C], Krueppel-like factor 5 (Basic transcription element-binding protein 2) (BTE-binding protein 2) (Colon krueppel-like factor) (GC-box-binding protein 2) (Intestinal-enriched krueppel-like factor) (Transcription factor BTEB2) SPECIES-INDEPENDENT COMPONENTS ------------------------------ A key focus of this analysis is identifying components NOT primarily driven by species differences - these represent conserved expression patterns shared across all three coral species and likely reflect fundamental biological processes. Number of species-independent components: 10 Number of species-driven components: 25 Top species-independent components (lowest cross-species variation): • Component 30: CV=0.243, Weight=750.9 • Component 9: CV=0.288, Weight=471.5 • Component 3: CV=0.344, Weight=561.5 • Component 11: CV=0.381, Weight=593.0 • Component 31: CV=0.414, Weight=1116.3 Functional Categories in Species-Independent Components: • organelle: 270 genes • catalytic activity: 139 genes • cytosol: 135 genes • nucleus: 129 genes • anatomical structure development: 124 genes • plasma membrane: 90 genes • molecular function regulator activity: 78 genes • hydrolase activity: 72 genes Top Biological Processes (conserved across species): • positive regulation of transcription by RNA polymerase II: 36 genes • cytoplasmic translation: 31 genes • translation: 30 genes • regulation of transcription by RNA polymerase II: 28 genes • negative regulation of transcription by RNA polymerase II: 27 genes • positive regulation of gene expression: 24 genes ANNOTATION STATISTICS --------------------- Across all components analyzed: • Total genes: 3525 • Annotated: 1334 (37.8%) • Unannotated: 2191 (62.2%) • Mean % unannotated per component: 62.2% • Components with >50% unannotated: 33 This indicates that a significant portion of the coral transcriptome remains functionally uncharacterized, highlighting the need for continued annotation efforts in non-model marine organisms. Species-Independent Components Annotation: • Mean % unannotated: 62.8% • Range: 43.6% - 76.0% CONCLUSIONS ----------- The tensor decomposition reveals that: 1. Gene expression patterns are largely conserved across the three coral species, suggesting shared stress response mechanisms and fundamental physiological processes. 2. Temporal dynamics show distinct phases, with some components showing early activation followed by decline, while others show progressive increases or stable expression throughout the experiment. 3. Species-specific expression signatures exist alongside conserved patterns, potentially reflecting evolutionary divergence in stress response strategies or ecological adaptations. 4. Key biological processes including transcriptional regulation, translation, signal transduction, and metabolic pathways are dynamically regulated across the experimental timeline. 5. Species-independent components reveal conserved cellular machinery including translation, metabolic regulation, and stress response pathways that are fundamental to coral biology regardless of species identity. 6. A substantial proportion of genes remain unannotated, indicating significant unexplored biology in coral genomes that warrants further investigation. PHYSIOLOGICAL CONTEXT --------------------- Integration with physiological measurements reveals the following patterns: • Chlorophyll content shows species-specific baseline differences, with dynamic changes across time points that may correlate with photosynthetic capacity and symbiont density. • Symbiodinium density varies between species, reflecting differences in symbiotic relationships and potentially heat tolerance strategies. • Host biomass (AFDW) shows temporal dynamics that may be linked to the metabolic and stress response gene expression patterns identified in the tensor decomposition. • Calcification rates differ between species, with potential links to components involved in biomineralization and ion transport gene expression. • Protein content dynamics across time points may reflect the translation and protein synthesis pathways identified as major themes in the gene expression components. These physiological measurements provide functional validation of the gene expression patterns and help contextualize the molecular stress response signatures across the three coral species. METHODS NOTE ------------ Analysis performed using barnacle tensor decomposition with rank-35 optimization and lambda_gene=0.2. Input data consisted of orthologous gene expression from 9800 ortholog groups across 30 samples and 4 time points. Visualizations and analysis performed using Python with matplotlib, seaborn, numpy, and pandas. GO term enrichment from UniProt annotations.