Data Card: DILGOM Whole-Blood Microarray Dataset (E-TABM-1036)

Summary

Expression + sample metadata + feature metadata for E-TABM-1036, which profile whole-blood gene expression in a population-based cohort from the Helsinki/Vantaa area of Finland. Profiled on Illumina HumanHT-12 v3 Expression BeadChip. This accession has no obesity or metabolic-syndrome phenotype data. These are the 'healthy' controls.

Source accession

Accession	N (SDRF rows)	Platform	Submitter
E-TABM-1036	518	Illumina HumanHT-12 v3	Inouye (Wellcome Trust Sanger Institute)

Files

• sample_metadata.parquet — one row per sample: sample_id, accession, age_band, sex. This is the full extent of usable per-sample metadata in the SDRF — see Notes.
• feature_metadata.parquet — one row per probe (IlluminaID): annotation from illuminaHumanv3.db (SYMBOL, ENTREZID, ENSEMBL, GENENAME, UNIPROT) plus reannotation QC fields from the package's internal ExtraInfo table (ProbeQuality, CodingZone, GenomicLocation, SecondMatches, OtherGenomicMatches, RepeatMask, OverlappingSNP, ProbeSequence)
• expression/ — partitioned parquet dataset (by accession, for schema consistency with other datasets in this collection even though there is only one accession here), long format: IlluminaID, sample_id, accession, value

Expression value processing (from the accession's IDF)

Per the submitter's own protocol description (bioassay_data_transformation protocol, P-TABM-5309):

"Normalized intensities were calculated via quantile normalization of the weighted-sums of the raw replicate intensities. Weighting was by number of beads contributing to raw signal. Final intensities were arrived at by a log2 transformation after quantile normalization."

RNA source: PAXgene Blood RNA Tubes, whole blood (per protocol P-TABM-5305), amplified via in vitro transcription (P-TABM-5306) and hybridized to HumanHT-12 v3 BeadChips (P-TABM-5307).

Provenance / reproducibility

Data retrieved with ArrayExpress::getAE() against BioStudies (EBI), with probe annotation from Bioconductor's illuminaHumanv3.db. See pull_dilgom.R