Evidence card: NCBI:GeneID:117690579

Real public-study evidence. Association evidence only; not a validated biomarker.

How to interpret expression as evidence for resilience. Differential expression alone does not establish a resilience marker; higher expression does not necessarily mean greater resilience.

This feature was selected among up to ten displayed findings in at least one study. All available rows for it are shown below, including nonsignificant and exploratory results. Presence in multiple datasets does not establish independent replication: exposures, tissues, life stages, and phenotype meanings differ. No cross-study score or pooled effect is calculated.

Guide to effect sizes, uncertainty and gene descriptions.

Study-specific effects

Study / design	RefSeq description	Contrast	Effect scale	Effect	Standard error	Adjusted p-value	Interpretation
CGIG_HEAT24H_RNASEQ_PRJNA407831	small nucleolar RNA U3	35C_24h_vs_0h_gill	log2 fold change	6.177	0.469	9.356e-36	Study-level association; see design caveats
CGIG_HEAT6H_RNASEQ_PRJNA407831	small nucleolar RNA U3	35C_6h_vs_0h_gill	log2 fold change	3.468	0.5415	8.225e-08	Study-level association; see design caveats
PRJEB86646	small nucleolar RNA U3	30C_vs_23C_fed_before_oshv1	log2 fold change	1.86	0.6411	0.01343	Study-level association; see design caveats
PRJNA516762	small nucleolar RNA U3	heat_35C_2h_vs_control_12C	log2 fold change	3.759	0.9701	0.02907	Study-level association; see design caveats
CGIG_HEAT_RNASEQ_PRJNA913164	small nucleolar RNA U3	30C_seawater_vs_20C_adult_ctenidium	log2 fold change	1.314	0.2581	7.41e-06	Study-level association; see design caveats
PRJNA298285	small nucleolar RNA U3	pH7.9_vs_pH8.1_20C_larvae	log2 fold change	0.4982	0.2463	0.9086	Study-level association; see design caveats
CGIG_MULTISTRESS_RNASEQ_PRJNA913164	small nucleolar RNA U3	30C_seawater_then_44C_emersion_6h_vs_20C_adult_ctenidium	log2 fold change	3.422	0.3256	4.323e-23	Study-level association; see design caveats
PRJNA826964	small nucleolar RNA U3	ocean_acidification_vs_ambient_control	log2 fold change	-0.2454	0.7532	0.959	Study-level association; see design caveats
PRJNA1329250	small nucleolar RNA U3	oshv1_microvariant_challenge_vs_unchallenged_control	log2 fold change	0.3604	0.3703	0.5409	Study-level association; see design caveats
PRJNA593309	small nucleolar RNA U3	oshv1_cohabitation_48h_vs_0h_21C	log2 fold change	2.821	0.9986	0.07232	Study-level association; see design caveats
CGIG_PESTO_F0_RNASEQ_PRJEB58194	small nucleolar RNA U3	pesticide_mixture_0_48hpf_vs_control_F0_gastrula_dlarva	log2 fold change	-0.4009	0.3723	0.649	Study-level association; see design caveats
CGIG_PESTO_F1DIRECT_RNASEQ_PRJEB58194	small nucleolar RNA U3	pesticide_mixture_TE_vs_TT_F1_gastrula_pediveliger	log2 fold change	-1.178	0.3851	0.3965	Study-level association; see design caveats
PRJNA756710	small nucleolar RNA U3	low_salinity_10ppt_vs_30ppt_during_vibrio_infection	log2 fold change	3.85	0.8459	0.0001975	Study-level association; see design caveats
PRJNA694496	small nucleolar RNA U3	thermotolerance_selected_vs_nonselected_control	log2 fold change	4.779	1.163	0.0002349	Study-level association; see design caveats
CGIG_TIRELEACHATE_RNASEQ_PRJNA856813	small nucleolar RNA U3	tire_leachate_high_vs_control_gill_day41	log2 fold change	0.2295	0.8404	0.9985	Study-level association; see design caveats
CGIG_TIREPARTICLE_RNASEQ_PRJNA856813	small nucleolar RNA U3	tire_microparticles_high_vs_control_gill_day41	log2 fold change	0.4554	0.9272	0.9999	Study-level association; see design caveats

CGIG_HEAT24H_RNASEQ_PRJNA407831

Phenotype interpretation: Intertidal-origin F1 oysters showed higher survival after acute heat stress (42 C, 1 h) and higher metabolic rates at 35 C than subtidal-origin oysters, with slower growth, and more transcriptional plasticity. AREE records the gill transcriptional response to 35 C averaged over the four populations, not survival per library.

Biological replication: 23 libraries analysed (24 sequenced; one 0 h outlier left out): 11 at 0 h and 12 at 24 h. Each library pools equal gill RNA from five oysters of one culture tank; four populations (two sites x intertidal or subtidal origin) x three tanks, and the same tanks are sampled at each time.

Evidence classification: stress_response. Tissue: gill; life stage: adult.

Quality flags: raw_reanalysis;acute_heat_response;pooled_libraries;population_blocked_contrast;shared_control_group;dependence_group=PRJNA407831_0h_pools;heat_24h;fdr_lt_0.05.

Study limitations:

Both AREE records from this study (the other is CGIG_HEAT6H_RNASEQ_PRJNA407831, 6 h) test against the same twelve 0 h pools, so they carry the dependence group PRJNA407831_0h_pools and count as one replication group.
Each library pools gill RNA of five oysters, so the error model describes tank-pool variation, not variation between animals. The sister IOCAS study PRJNA826964 deposited resequenced RNA as replicates, so these pools were checked with the expressed-SNP test (scripts/check_replicate_genotypes.py) before analysis.
The model blocks on the four source populations (~ population + condition), so the effect is averaged over intertidal and subtidal origins and the two sites; the publication’s main result, that intertidal oysters respond with more plasticity, is not part of this record. A tank-paired model (~ tank + condition) is kept as a sensitivity analysis.
Sample aliases code tidal origin as J or X (for example BYQJ01, LTX243); the publication uses I (intertidal) and S (subtidal), and the BioSamples do not record it. J and X are read as the Chinese jian (intertidal) and xia (subtidal), but no AREE result depends on that reading because population is a blocking factor.
One 0 h library, h0_LTJ_2, is left out by AREE’s QC rule (median VST correlation below the median minus three robust SDs); it is the library the publication also discarded (LT-I_0h_2). The exclusion lives in scripts/prepare_prjna407831_fastqs.py.
ENA served other studies at 0.05-0.35 MB/s in October 2026, so all runs come from NCBI .sra archives checked against NCBI’s MD5s, with read counts checked against the manifest on conversion.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: ba65a504fd1c43fe5df5636afd198fc3f2b95abd6a145604c6ed9ca3caf03421.

CGIG_HEAT6H_RNASEQ_PRJNA407831

Biological replication: 23 libraries analysed (24 sequenced; one 0 h outlier left out): 11 at 0 h and 12 at 6 h. Each library pools equal gill RNA from five oysters of one culture tank; four populations (two sites x intertidal or subtidal origin) x three tanks, and the same tanks are sampled at each time.

Evidence classification: stress_response. Tissue: gill; life stage: adult.

Quality flags: raw_reanalysis;acute_heat_response;pooled_libraries;population_blocked_contrast;shared_control_group;dependence_group=PRJNA407831_0h_pools;heat_6h;fdr_lt_0.05.

Study limitations:

Both AREE records from this study (the other is CGIG_HEAT24H_RNASEQ_PRJNA407831, 24 h) test against the same twelve 0 h pools, so they carry the dependence group PRJNA407831_0h_pools and count as one replication group.
Each library pools gill RNA of five oysters, so the error model describes tank-pool variation, not variation between animals. The sister IOCAS study PRJNA826964 deposited resequenced RNA as replicates, so these pools were checked with the expressed-SNP test (scripts/check_replicate_genotypes.py) before analysis.
The model blocks on the four source populations (~ population + condition), so the effect is averaged over intertidal and subtidal origins and the two sites; the publication’s main result, that intertidal oysters respond with more plasticity, is not part of this record. A tank-paired model (~ tank + condition) is kept as a sensitivity analysis.
Sample aliases code tidal origin as J or X (for example BYQJ01, LTX243); the publication uses I (intertidal) and S (subtidal), and the BioSamples do not record it. J and X are read as the Chinese jian (intertidal) and xia (subtidal), but no AREE result depends on that reading because population is a blocking factor.
One 0 h library, h0_LTJ_2, is left out by AREE’s QC rule (median VST correlation below the median minus three robust SDs); it is the library the publication also discarded (LT-I_0h_2). The exclusion lives in scripts/prepare_prjna407831_fastqs.py.
ENA served other studies at 0.05-0.35 MB/s in October 2026, so all runs come from NCBI .sra archives checked against NCBI’s MD5s, with read counts checked against the manifest on conversion.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: 3ff5b26926b5dc5ac2b94196d5b5a8628742dc7a4ebf2460da80389567311c2a.

CGIG_HEAT_RNASEQ_PRJEB86646

Phenotype interpretation: Lower mortality after OsHV-1 mu-Var cohabitation is favorable. The publication reports that 30 C limits POMS development and improves survival, through metabolic reprogramming and stronger baseline and induced antiviral immunity. AREE records the whole-spat transcriptional state at 30 C before infection, not survival per library.

Biological replication: 12 RNA-seq libraries, all sampled at 0 h post-cohabitation: 6 at 23 C and 6 at 30 C, three per temperature in each of the two families. Each library pools the soft tissue of 10 live spat taken from one tank, as in the DECICOMP starvation arm.

Evidence classification: resilience_associated. Tissue: whole_soft_tissue; life stage: spat.

Quality flags: raw_reanalysis;temperature_response;pre_infection_baseline;family_blocked_contrast;shared_control_group_with_PRJEB86618;fdr_lt_0.05;dependence_group=DECICOMP_PRJEB86525_0h_controls.

Study limitations:

The six 23 C fed control libraries (PRJEB86525, 0 h) are the same libraries used as controls in CGIG_STARVATION_RNASEQ_PRJEB86618. The two harmonized studies therefore share a reference group, so their effect estimates are correlated and must not be treated as independent in a meta-analysis.
The contrast uses only the 0 h libraries, sampled when the OsHV-1 mu-Var donors were added, so it captures temperature without infection. The 36 later libraries (3-72 h) follow infection, which progresses far less at 30 C, and are not used.
Each tank held both families, so the F11N and F14R libraries with the same replicate number probably share a tank. The model ~ family + condition treats the six libraries per temperature as independent, although only three tanks per temperature are independent units.
The 30 C arm and the 23 C controls were held in separate tanks of one experiment at the Argenton station, so tank is nested in temperature.
Condition, family, time and replicate come from the ENA sample aliases and titles (sam_P_<3 fed 23 C | 4 fed 30 C>_), and the two sources agree for every library.
The publication reports enhanced baseline antiviral immunity at 30 C. At 0 h this reanalysis finds the opposite for several classic antiviral markers: all 13 significant interferon-induced protein 44 genes, 10 of 12 significant tumor necrosis factor genes and both significant 2’-5’-oligoadenylate synthetase genes are lower at 30 C. The publication is not open access and its gene lists were not compared, so whether this reflects different gene sets, a different time point or a real disagreement is unresolved.
Pools at 30 C map about three points less to the transcriptome than pools at 23 C. The shift is systematic but small beside the expression effects.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: abbc33a2dc6e48a0f0a903bbca1d4bce3f90c56927e80ff5c5bc3e88dd6c0c5c.

CGIG_HEAT_RNASEQ_PRJNA516762

Phenotype interpretation: Organism-level thermal-tolerance direction was not measured; molecular heat-response direction only

Biological replication: Direct heat-versus-control contrast: 3 pooled RNA-seq libraries per condition, each pooling 5 individuals (30 individuals total; 6 libraries)

Evidence classification: stress_response. Tissue: gill; life stage: juvenile.

Quality flags: raw_reanalysis;acute_heat_response;pooled_libraries;fdr_lt_0.05.

Study limitations:

This intake represents molecular heat-stress response, not a direct survival or organism-level thermal-tolerance measurement.
Each RNA-seq library pooled gill RNA from five oysters, so individual-level variability is unavailable.
AREE independently reanalyzed the raw reads against GCF_963853765.1/RS_2024_06; the original study used legacy assembly GCA_000297895.1 and CGI/newGene identifiers, so these results are a current-reference reanalysis rather than an exact reconstruction of the publication pipeline.
The direct heat-versus-control contrast yields modest differential expression (99 genes at FDR <= 0.05 of 22,397 tested), and PC1 (26.3% of variance) separates most but not all libraries, with heat_A overlapping the controls.
The HSF1 RNAi-only and RNAi-plus-heat contrasts are preserved in the source workbook but excluded from this direct heat-versus-control reanalysis.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: f0bded9f7662170c4985ff78b455a69e7df0b43c114895da85277159a2daba88.

CGIG_HEAT_RNASEQ_PRJNA913164

Phenotype interpretation: Mortality was elevated in the stress treatments relative to the control. AREE records the ctenidium transcriptional response to elevated seawater temperature, not survival per library.

Biological replication: 44 ctenidium libraries analysed, one oyster each (23 controls and 21 treated after QC exclusions), from 48 sequenced: 24 controls and 24 30 C, 12 diploid and 12 triploid in each arm. The 24 control libraries are shared with CGIG_MULTISTRESS_RNASEQ_PRJNA913164 (30 C followed by 44 C emersion).

Evidence classification: stress_response. Tissue: gill; life stage: adult.

Quality flags: raw_reanalysis;quantseq_3prime_tag_seq;ploidy_blocked_contrast;shared_control_group;dependence_group=PRJNA913164_20C_controls;temperature_response;fdr_lt_0.05.

Study limitations:

The BioSamples carry no treatment or ploidy attribute. Both come from the description column of the authors’ NCBI submission sheet (github.com/mattgeorgephd/NOPP-gigas-ploidy-temp, 202107_EXP2/tag-seq/NCBI/attributes.tsv), matched to BioSample by isolate name; sample-name prefixes do not map one-to-one to groups.
The study has three arms that share one control group. AREE records each stressor against the 20 C controls as its own study (CGIG_MULTISTRESS_RNASEQ_PRJNA913164 (30 C followed by 44 C emersion) is the other), so the two records share their reference group and their effect estimates are correlated; they must not be treated as independent in a meta-analysis.
Diploids and triploids are pooled with ploidy as a blocking factor (~ ploidy + condition), so the effect is averaged over ploidy. The publication’s main finding is that triploids respond differently, especially to the multiple stressor; the ploidy-by-treatment interaction is reported in the quality-control summary, and ploidy-specific effects are not part of this record.
QuantSeq 3’ tag-seq yields about one read per transcript, so Salmon was run with –noLengthCorrection and DESeq2 used the counts without the tximport length offset (–tag-seq=yes). Library sizes are small (0.9-8.3 million reads), and isoform-level information is not recoverable.
The authors excluded eight libraries as outliers (D54, N56, X44, R53, M43, N54, X42, T62). AREE quantified all 72 and applied its own rule (median VST correlation below the median minus three robust SDs): seven of the authors’ eight fail it, D54 passes and is kept, and R62 and N57 also fail and are left out. The exclusions live in scripts/prepare_prjna913164_fastqs.py and remove one or two libraries from some ploidy-by-treatment cells.
The expressed-SNP replicate check (scripts/check_replicate_genotypes.py) could not be run: 3’ tags cover too little of the target transcripts (about 2,000 usable reads per library, even after poly-A and adapter trimming) to call allele frequencies. Each library is a single adult oyster according to the publication, so no pooling or resequencing is expected, but this was not verified genetically.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: f77e0e6f5683b7fe006272a6c53cb16bb293cf3542164602272e155aa0f4bafb.

CGIG_LARVAL_OA_RNASEQ_PRJNA298285

Phenotype interpretation: No phenotype is deposited and no publication is linked; AREE records the larval transcriptional response averaged over three developmental stages.

Biological replication: 12 whole-larva libraries: 6 ambient and 6 acidified (pH 7.9, 20 C), two per developmental stage. Each library is presumably a pool of many larvae; whether the two libraries per cell are separate cultures or one culture is not recorded and is tested with the expressed-SNP check.

Evidence classification: environmental_exposure. Tissue: whole_organism; life stage: larva.

Quality flags: raw_reanalysis;larval_development;pooled_libraries;stage_blocked_contrast;shared_control_group;dependence_group=PRJNA298285_ambient_larvae;no_linked_publication;ocean_acidification_response;replicate_culture_independence_unverified.

Study limitations:

No publication is linked to the BioProject, and searches of PubMed and Europe PMC (including the accessions PRJNA298285 and SRP064660) on 2026-10-07 found none, so there are no published methods, carbonate chemistry, survival data or results to reconcile against.
The three AREE records from this study (acidified, warmed, acidified + warmed) test against the same six ambient libraries, so they carry the dependence group PRJNA298285_ambient_larvae and count as one replication group.
There are two libraries per stage and treatment. The stage block (~ stage + condition) treats them as replicates; whether they come from separate cultures is not recorded and is checked with scripts/check_replicate_genotypes.py.
The treatments are mild (pH 7.9 against 8.1; 22 C against 20 C), so small transcriptional effects are expected.
Treatment, stage and replicate come from the ENA sample aliases (C.gigas_) and library names (<Trc | Vg1 | Vg2>_<ambi | acid | warm | intr><A | B>), and agree with the BioSample pH, temperature, dev_stage and age attributes. One library name is a file name (C3759_TAGCTT_L003_R1_001.fastq.gz, late-veliger warmed); it is taken as replicate B because the other late-veliger warmed library is C3752_Vg2_warmA. One library (D5236_Trc_warmA) is from a later sequencing batch at 126 bp.
All runs come from NCBI .sra archives checked against NCBI’s MD5s, with read counts checked against the manifest on conversion.
The A and B libraries of each cell have nearly identical expression profiles and cannot be distinguished genetically from libraries of other treatments, so they may be subsamples of one culture; if so, the error model understates variability and significance is optimistic. The record carries the flag replicate_culture_independence_unverified.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: 2bae06ceaff293aaaa0e3dbf9ec6cce42ea0b18c4028ba6ef4e2a374cfeee32b.

CGIG_MULTISTRESS_RNASEQ_PRJNA913164

Phenotype interpretation: Mortality after the multiple stressor was higher in triploids than diploids (36.4% versus 14.8%), with metabolic depression and reduced ctenidium Na/K-ATPase activity in triploids. AREE records the ctenidium transcriptional response averaged over ploidy, not survival per library.

Biological replication: 42 ctenidium libraries analysed, one oyster each (23 controls and 19 treated after QC exclusions), from 48 sequenced: 24 controls and 24 30 C + 44 C emersion, 12 diploid and 12 triploid in each arm. The 24 control libraries are shared with CGIG_HEAT_RNASEQ_PRJNA913164 (30 C seawater only).

Evidence classification: stress_response. Tissue: gill; life stage: adult.

Quality flags: raw_reanalysis;quantseq_3prime_tag_seq;ploidy_blocked_contrast;shared_control_group;dependence_group=PRJNA913164_20C_controls;multiple_stressor_response;fdr_lt_0.05.

Study limitations:

The BioSamples carry no treatment or ploidy attribute. Both come from the description column of the authors’ NCBI submission sheet (github.com/mattgeorgephd/NOPP-gigas-ploidy-temp, 202107_EXP2/tag-seq/NCBI/attributes.tsv), matched to BioSample by isolate name; sample-name prefixes do not map one-to-one to groups.
The study has three arms that share one control group. AREE records each stressor against the 20 C controls as its own study (CGIG_HEAT_RNASEQ_PRJNA913164 (30 C seawater only) is the other), so the two records share their reference group and their effect estimates are correlated; they must not be treated as independent in a meta-analysis.
Diploids and triploids are pooled with ploidy as a blocking factor (~ ploidy + condition), so the effect is averaged over ploidy. The publication’s main finding is that triploids respond differently, especially to the multiple stressor; the ploidy-by-treatment interaction is reported in the quality-control summary, and ploidy-specific effects are not part of this record.
QuantSeq 3’ tag-seq yields about one read per transcript, so Salmon was run with –noLengthCorrection and DESeq2 used the counts without the tximport length offset (–tag-seq=yes). Library sizes are small (0.9-8.3 million reads), and isoform-level information is not recoverable.
The authors excluded eight libraries as outliers (D54, N56, X44, R53, M43, N54, X42, T62). AREE quantified all 72 and applied its own rule (median VST correlation below the median minus three robust SDs): seven of the authors’ eight fail it, D54 passes and is kept, and R62 and N57 also fail and are left out. The exclusions live in scripts/prepare_prjna913164_fastqs.py and remove one or two libraries from some ploidy-by-treatment cells.
The expressed-SNP replicate check (scripts/check_replicate_genotypes.py) could not be run: 3’ tags cover too little of the target transcripts (about 2,000 usable reads per library, even after poly-A and adapter trimming) to call allele frequencies. Each library is a single adult oyster according to the publication, so no pooling or resequencing is expected, but this was not verified genetically.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: d2d6e4a07aa73594d97a4c9c4a593cace73fecddc5c11304a2bbef65277282d6.

CGIG_OA_RNASEQ_PRJNA826964

Phenotype interpretation: The BioProject title frames acidification as compromising energy management; AREE records the hepatopancreas transcriptional response to acidification, not a measured organism-level energetic or survival phenotype

Biological replication: 6 pooled RNA samples, one per arm at each of 7, 28 and 56 days, each sequenced as three libraries (18 libraries); the BioSample records call the libraries biological replicates, but their genotypes show each set of three is one pool of several oysters

Evidence classification: stress_response. Tissue: hepatopancreas; life stage: not_recorded.

Quality flags: raw_reanalysis;ocean_acidification_response;timepoint_blocked_contrast;pooled_rna_samples;possible_genetic_background_confound.

Study limitations:

No publication is linked to the BioProject, and searches of PubMed, Europe PMC full text and Crossref on 2026-09-28 found none, so there are no published methods, results or effect tables to reconcile the reanalysis against.
The acidification intensity (target pH or pCO2), seawater carbonate chemistry, temperature and oyster size or age are not recorded in the public metadata, so exposure intensity and life stage are unresolved.
The treatment was probably CO2-driven acidification to pH 7.50. This is inferred, not documented: the same IOCAS group (L. Qiu and colleagues) ran C. gigas CO2 exposures at pH 7.50 sampled from 7 to 56 days (Wang et al. 2021, Ecotoxicol Environ Saf, doi:10.1016/j.ecoenv.2021.112235) and at pH 7.50 for 7 and 60 days (Wang et al. 2020, Sci Total Environ, doi:10.1016/j.scitotenv.2020.140177), and reported hepatopancreas responses to acidification over about two months (Zhang et al. 2021, J Oceanol Limnol, doi:10.1007/s00343-021-0462-x). None of these papers is linked to PRJNA826964, and the closed-access hepatopancreas paper, the most likely companion, has not been checked against it.
The primary AREE contrast compares the three acidified pools with the three control pools and blocks on sampling day (~ timepoint + condition); it estimates an acidification effect averaged over 7, 28 and 56 days rather than time-specific or acclimation effects, and it has only two residual degrees of freedom.
The three libraries per day-by-condition cell are technical replicates of one pooled RNA sample, not the independent biological replicates the BioSample attributes state. Expressed-SNP allele frequencies (scripts/check_replicate_genotypes.py, 4M read pairs per library) agree within a cell to within resequencing noise and show no single-animal genotype peaks, and within-cell expression dispersion is 2.7% BCV against 10.8-22.4% in the other AREE oyster studies. Treating the 18 libraries as replicates inflates significance roughly twelvefold, so the libraries are summed into 6 pools (–collapse-by=condition+timepoint).
Each pool is a different set of animals, and the acidified pools are genetically further from the controls (mean allele-frequency difference 0.24-0.31) than the control pools are from one another (0.085-0.19), so acidified and control oysters may come from different stocks or families. Several of the strongest effects are 4-6 log2 changes in transposon-derived genes (tigger, HARBI1), the pattern expected from presence-absence variation between genetic backgrounds, so part of the contrast may be genetic rather than an acidification response. The evidence is harmonized with every row flagged pooled_rna_samples and possible_genetic_background_confound, so scoring discounts it and evidence cards list the caveat; treat it as provisional acidification evidence until a design with shared genetic backgrounds confirms it.
Condition, sampling day and replicate were taken from NCBI BioSample treatment and replicate attributes and agree with the library names (C/O = control/acidified, followed by day and replicate).
The BioProject carries no author-deposited processed expression matrix.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: 4a01845d837665c4e977190e7e9111a2d5a7bd3e5f38d89ca473276a96b26f2c.

CGIG_OSHV1_RNASEQ_PRJNA1329250

Phenotype interpretation: Infection up-regulated antiviral defense transcripts, including RIG-I-like receptors and the interferon-like pathway, similarly in both lineages; AREE records molecular antiviral responsiveness, not an organism-level survival difference

Biological replication: 42 RNA-seq libraries: 12 controls (6 Midori, 6 Miyagi) and 30 OsHV-1 microvariant-challenged (15 Midori, 15 Miyagi, split across USA, Australia, and France viral strains at 5 libraries per lineage-strain cell)

Evidence classification: disease_associated. Tissue: whole_soft_tissue; life stage: spat.

Quality flags: raw_reanalysis;pathogen_challenge_response;lineage_balanced_contrast.

Study limitations:

This intake represents the molecular antiviral response to experimental OsHV-1 microvariant challenge, not a measured survival or organism-level disease-resistance difference between the two lineages.
Challenge dose, exposure duration, and time of sampling after inoculation are not recorded in the public SRA metadata, so exposure intensity and duration are unresolved.
The challenged and control arms are unequal in size (30 versus 12 libraries), but lineage is balanced within each arm and within every viral strain: 6 Midori and 6 Miyagi controls, and 5 of each lineage for each of the Australia, France, and USA strains. The lineage effect is therefore not confounded with infection status, and AREE blocks on lineage as ~ population + condition.
The publication contrasts the USA and Australia microvariant strains and reports no marked difference between them; the France-strain libraries are present in the BioProject and are pooled into the challenged arm of the AREE contrast rather than analyzed as a separate strain effect.
The BioProject carries no author-deposited processed expression matrix, so AREE has no published feature-level effect table to reconcile against.
Condition, lineage, and viral-strain assignments were reconstructed from NCBI BioSample attribute tags rather than from a supplementary sample table; the source metadata spells the Miyagi lineage ‘Myagi’, which AREE normalized.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: 3beb15844b8040ff58c64acd452381f853bbbacb730cce2bf434b6ee05d41f2e.

CGIG_OSHV1_RNASEQ_PRJNA593309

Phenotype interpretation: At 21 C survival 14 days after cohabitation was 52.4%, and OsHV-1 DNA had reached high levels in all three oysters sampled at 48 h. AREE records the whole-oyster transcriptional response to infection, not survival per library.

Biological replication: 6 libraries, one whole 8-month-old oyster per tank: three at 0 h and three at 48 h, the same three 21 C tanks sampled at both times.

Evidence classification: disease_associated. Tissue: whole_organism; life stage: juvenile.

Quality flags: raw_reanalysis;pathogen_challenge_response;unpaired_primary_tank_paired_sensitivity;one_oyster_per_tank;dependence_group=PRJNA593309_21C_tanks.

Study limitations:

The 0 h and 48 h oysters come from the same three tanks. The primary model is unpaired (~ condition); a tank-paired model (~ tank + condition) is kept as a sensitivity analysis.
One whole oyster per tank was sequenced at each time point, with three replicate tanks per temperature (45 l tanks), so libraries are tank-level replicates and the contrast has three tanks per arm.
The two AREE records from this study use different oysters but the same three 21 C tanks (the 21 C 12 h oysters of CGIG_HEAT_RNASEQ_PRJNA593309 and the 21 C 0 h and 48 h oysters of CGIG_OSHV1_RNASEQ_PRJNA593309), so both carry the dependence group PRJNA593309_21C_tanks and count as one replication group.
The BioSamples give a collection date of 2019-05-19, but the publication dates the experiment to May 2016; the experimental groups follow the sample aliases (Cg_<CTR | OsHV>hpc_), which agree with the publication’s design.
The study also deposited libraries at 26 C and at 24 and 96 h; the publication does not analyse the 26 C libraries, and AREE uses only the time points of its two contrasts.
The publication aligned reads with STAR to the Ensembl oyster genome v9.38 and tested with edgeR; AREE quantifies against GCF_963853765.1/RS_2024_06 with Salmon and DESeq2.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: cca3b26b0a11f8cd29467b1df45d728788247bf18855e27becda42e3e88e27d8.

CGIG_PESTO_F0_RNASEQ_PRJEB58194

Phenotype interpretation: The publication found no difference in embryo-larval abnormalities at 24 hpf, disturbed swimming, larger and heavier exposed larvae at some times, a lower epinephrine-induced metamorphosis rate and higher field survival, and differential expression of developmental genes (Hox orthologues, cell-cycle regulators) in gastrulae. AREE records the transcriptional response averaged over two stages.

Biological replication: 36 libraries: 18 control and 18 exposed, nine per stage (gastrula, D larva), labelled with group digits 1-3 and replicate letters A-F.

Evidence classification: environmental_exposure. Tissue: whole_organism; life stage: embryo_larva.

Quality flags: raw_reanalysis;pollutant_exposure_response;pesticide_mixture;pooled_libraries;stage_blocked_contrast;direct_exposure.

Study limitations:

One control D-larva sample title (BT2Ed) is attached to two runs (ERR10659105, ERR10659106); both are kept with the run accession appended to the replicate, and the expressed-SNP check tests whether they are one sample.
The exposure is a mixture of 18 pesticides and metals at environmentally relevant concentration (nominal sum 2.85 ug/L; measured 3.74 ug/L in F0) during embryo-larval development (0-48 h post fertilization) at the Ifremer Bouin hatchery.
Each library is presumably a pool of many embryos or larvae; pool size, culture vessels and the meaning of the F0 group digits (1-3) are not recorded in the public metadata, and pools were checked with the expressed-SNP test (scripts/check_replicate_genotypes.py) before analysis.
All runs come from NCBI .sra archives checked against NCBI’s MD5s (11 F0 runs have no ENA FASTQ), with read counts checked against the NCBI spot counts on conversion.
The study also deposited F0 pediveligers (sampled after exposure) and F1 offspring exposed in both generations (EE); AREE does not use them.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: 5eb3fa1130b516be240eecebcd8c88a5b405a6739f99d273764f3054814af5bb.

CGIG_PESTO_F1DIRECT_RNASEQ_PRJEB58194

Phenotype interpretation: The publication found that the parents’ exposure shaped the offspring phenotype more than direct exposure did; offspring of exposed parents had lower epinephrine-induced metamorphosis and field survival. AREE records the transcriptional response to direct exposure averaged over two stages.

Biological replication: 12 libraries: 6 TT and 6 TE, three per stage.

Evidence classification: environmental_exposure. Tissue: whole_organism; life stage: embryo_larva.

Quality flags: raw_reanalysis;pollutant_exposure_response;pesticide_mixture;pooled_libraries;stage_blocked_contrast;shared_control_group;dependence_group=PRJEB58194_F1_TT;direct_exposure.

Study limitations:

The two F1 records (direct exposure TE and parental exposure ET) test against the same six TT libraries, so they carry the dependence group PRJEB58194_F1_TT; only the direct-exposure record is harmonized.
The exposure is a mixture of 18 pesticides and metals at environmentally relevant concentration (nominal sum 2.85 ug/L; measured 3.74 ug/L in F0) during embryo-larval development (0-48 h post fertilization) at the Ifremer Bouin hatchery.
Each library is presumably a pool of many embryos or larvae; pool size, culture vessels and the meaning of the F0 group digits (1-3) are not recorded in the public metadata, and pools were checked with the expressed-SNP test (scripts/check_replicate_genotypes.py) before analysis.
All runs come from NCBI .sra archives checked against NCBI’s MD5s (11 F0 runs have no ENA FASTQ), with read counts checked against the NCBI spot counts on conversion.
The study also deposited F0 pediveligers (sampled after exposure) and F1 offspring exposed in both generations (EE); AREE does not use them.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: eeb95c68cdb8cacd7c256a0616f24042aaa54da88a33682cf82f476119e25b62.

CGIG_SALINITY_RNASEQ_PRJNA756710

Phenotype interpretation: Higher mortality after Vibrio infection is unfavorable; infected oysters reached 100% mortality at 10 ppt versus 33% at 30 ppt over 12 days (log-rank p < 0.01, n = 30 per tank in triplicate). AREE records the gill transcriptional response to low salinity during infection, not survival per library.

Biological replication: 12 RNA-seq libraries used for the contrast: 6 at 30 ppt and 6 at 10 ppt, three per salinity at each of 12 h and 48 h. Each library pools equal gill tissue from three oysters taken from one tank. Two further 0 h libraries (SRR15559848, SRR15559847) are shared pre-transfer baselines with no salinity contrast and are not used.

Evidence classification: disease_associated. Tissue: gill; life stage: not_recorded.

Quality flags: raw_reanalysis;low_salinity_response;infection_background_all_arms;timepoint_blocked_contrast;fdr_lt_0.05.

Study limitations:

Every sequenced oyster was infected with Vibrio alginolyticus, so the contrast estimates the effect of low salinity during infection, not of low salinity alone. The uninfected, sham-injected controls held at 30 ppt were used for the survival curve only and were not sequenced. stressor_class is salinity because salinity is the only factor that differs between the contrasted arms.
The primary AREE contrast pools 12 h and 48 h and blocks on sampling time (~ timepoint + condition). It estimates a low-salinity effect averaged over the two times, whereas the publication tested each time separately; each time-by-salinity cell has only three libraries.
Each salinity arm used three tanks, and a pool of three oysters was taken from each tank at each time, so the 12 h and 48 h libraries with the same replicate number may share a tank. The public metadata do not state that replicate numbers correspond to tanks, so the model does not pair them.
Oyster age or size class is not given beyond an average wet weight of 18.2 +/- 2.8 g, so life_stage is not recorded. Temperature during the challenge is not reported.
The library names use C23 for the normal-salinity arm, but the publication and BioSample titles give the normal salinity as 30 ppt; C10 matches the 10 ppt arm.
Condition, time and replicate come from the NCBI BioSample titles. They agree with the library names (C10/C23 followed by 1-3 for 12 h and 4-6 for 48 h), and the raw read count of every library matches publication Supplementary Table 3.
The publication aligned reads with HISAT2 to GCA_902806645.1, counted them with featureCounts and ran DESeq2 per time point (padj < 0.05, |fold change| > 1). It reports 3,038 DEGs at 12 h (1,731 higher, 1,307 lower at 10 ppt) and 3,187 at 48 h (1,656 higher, 1,531 lower) in Supplementary Table 4, keyed by RefSeq LOC identifiers. Only significant genes are listed, so the table cannot be used as a full ranked result.
The companion 16S rRNA amplicon data for the digestive gland are in the separate BioProject PRJNA756403 and are outside this record.
Against the publication’s per-time DEGs (Supplementary Table 4), 5,206 of 6,225 LOC identifiers map directly to a current GeneID. Of those tested here, 96.0% (12 h) and 95.7% (48 h) change in the same direction in the pooled AREE model, fold changes correlate at r = 0.784 and 0.759, and none is significant in the opposite direction; 44% reach FDR < 0.05 in the pooled model.
rRNA genes are over-represented among low-salinity-higher genes (10 observed versus 0.8 expected; adjusted p = 1.2e-10), which points to a difference in residual rRNA between arms rather than a biological response; those loci should not be interpreted.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: 39f1098bfe2737737332f9e0cd5fdcd64159c848ac8f8562e1d8912615e59fe9.

CGIG_THERMOTOL_RNASEQ_PRJNA694496

Phenotype interpretation: Selected population had higher post-heat-shock survival (78.05% versus 66.34%) and field summer survival (73% versus 62%)

Biological replication: RNA-seq: 3 selected-population and 3 control-population larval libraries; number of larvae pooled per library was not reported

Evidence classification: resilience_associated. Tissue: whole_larva; life stage: umbo_larva.

Quality flags: raw_reanalysis;qc_reviewed;population_selection_association;constitutive_expression;selected_A_pc2_displacement;sensitivity_direction_robust;fdr_lt_0.05.

Study limitations:

The RNA-seq comparison is constitutive selected-versus-control expression at the larval stage, not an acute heat-versus-control contrast.
Thermotolerance survival was measured in related later-stage cohorts rather than the RNA-sequenced larval pools.
The publication reports 4,764 DEGs but does not provide the complete feature-level effect, p-value, or standard-error table in its supplementary archive.
The publication describes three replicates per group but does not report how many larvae were pooled into each RNA-seq library.
All six runs share one BioSample record, and its tissue field says gill despite the publication consistently describing whole umbo larvae; AREE follows the publication for tissue and records the discrepancy.
The source genome assembly and differential-expression software/threshold are not reported clearly enough for exact reproduction; AREE used GCF_963853765.1/RS_2024_06, so these results are a current-reference reanalysis rather than an exact reconstruction of the publication pipeline.
Selected_B and selected_C have lower Salmon mapping rates than the other libraries, and selected_A is displaced on PC2; sensitivity analysis supports the primary effect directions, but these patterns remain QC caveats.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: 636f73eb2b9094e50a8c3c3213792513bff4e26fc203a607471190188445b6a3.

CGIG_TIRELEACHATE_RNASEQ_PRJNA856813

Phenotype interpretation: The publication reports activation of energy metabolism and stress responses under leachate exposure, signs of microbiota dysbiosis at high concentrations, reduced sperm motility and oocyte transcriptional changes, but no significant ecophysiological change in adults and no effect on offspring growth, which it interprets as possible resilience. AREE records the gill transcriptional response, not a measured phenotype per library.

Biological replication: 12 day-41 gill libraries, one oyster each: 6 controls and 6 high-concentration leachate oysters.

Evidence classification: stress_response. Tissue: gill; life stage: adult.

Quality flags: raw_reanalysis;pollutant_exposure_response;shared_control_group;dependence_group=PRJNA856813_gill_controls;tire_leachate.

Study limitations:

The publication is not open access and was not read in full; exposure concentrations, tank layout and the number of tanks per treatment are therefore not recorded here. If each treatment was held in a single tank, oyster-level replicates would not account for tank clustering.
Both AREE records from this study test against the same six day-41 gill control oysters, so they carry the dependence group PRJNA856813_gill_controls and count as one replication group.
AREE uses only the day-41 gill libraries of the control, high-concentration leachate and high-concentration tire microparticle groups. The deposited study also has low concentrations, a natural-particle control, digestive gland (day 44), gonad (day 44, four per group), time-zero baselines and 16S amplicon data, which are not used.
Groups come from the SRA sample names (D41_GI_<CTL | LEA_H | MR_H>_; D41 = day 41, GI = gill); the BioSamples record the tissue as gills. The ENA portal was unavailable when the study was staged, so the reads come from NCBI .sra archives checked against NCBI’s MD5s, and each converted FASTQ must match the SRA spot count.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: 9f52f19bfaa110f7d5ba09e6d53e20073cd2ac0f540b2270c2aef8363909af42.

CGIG_TIREPARTICLE_RNASEQ_PRJNA856813

Biological replication: 12 day-41 gill libraries, one oyster each: 6 controls and 6 high-concentration tire-particle oysters.

Evidence classification: stress_response. Tissue: gill; life stage: adult.

Quality flags: raw_reanalysis;pollutant_exposure_response;shared_control_group;dependence_group=PRJNA856813_gill_controls;tire_microparticles.

Study limitations:

The publication is not open access and was not read in full; exposure concentrations, tank layout and the number of tanks per treatment are therefore not recorded here. If each treatment was held in a single tank, oyster-level replicates would not account for tank clustering.
Both AREE records from this study test against the same six day-41 gill control oysters, so they carry the dependence group PRJNA856813_gill_controls and count as one replication group.
AREE uses only the day-41 gill libraries of the control, high-concentration leachate and high-concentration tire microparticle groups. The deposited study also has low concentrations, a natural-particle control, digestive gland (day 44), gonad (day 44, four per group), time-zero baselines and 16S amplicon data, which are not used.
Groups come from the SRA sample names (D41_GI_<CTL | LEA_H | MR_H>_; D41 = day 41, GI = gill); the BioSamples record the tissue as gills. The ENA portal was unavailable when the study was staged, so the reads come from NCBI .sra archives checked against NCBI’s MD5s, and each converted FASTQ must match the SRA spot count.

Provenance: Reference GCF_963853765.1 / RS_2024_06; mapping exact; workflow 0.1.0.

Harmonized evidence. Processed input. Input SHA-256: a4d40ba861f26ac5d8a0c986666287ba5ad8ba6b4e7fd4766c34c1cf11c4e483.

Next validation step

Independent biological replication with matched exposure and measured phenotype, an appropriate experimental-unit model, and targeted validation in the relevant tissue and life stage.