SAIGE-QTL: Single-nucleus Expression QTL Summary Statistics using Human Brain Cohort

https://github.com/RajLabMSSM/SingleBrain

Full associations and top association summary statistics for cis-eQTLs mapped in the human brain single-nucleus RNA seq cohort(Fujita et al.[PMID: 38514782]) using SAIGE-QTL[PMID: 38798318], as part of the "SingleBrain" project.

Sample size = 402 European ancestry donors.

Each file has the following naming convention:

{Cell type}_eqtl_{ASSOC}.tsv.gz

References

The following reference was used for mapping phenotypes:

1. GENCODE - GENCODE v38 comprehensive transcripts (https://www.gencodegenes.org/human/release_38.html)

Cell type

The following brain 7 major cell types were tested for genetic association:

• Ast: astrocytes
• End: endothelial cells
• Ext: excitatory neurons
• IN: inhibitory neurons
• MG: microglia
• OD: oligodendrocytes
• OPC: oligodendrocyte progenitor cell

All phenotype matrices were scaled and centered and then quantile normalized.

Associations

Top associations (top_assoc.tsv.gz) list the SNP-feature pair with the lowest adjusted P-value (ACAT_P) for that feature.

Full associations (full_assoc.tsv.gz) list all tested SNP-feature pairs.

Data dictionary

The columns of the two association files only differ by the presence of the ACAT_P column in the top associations.

feature: the phenotype being tested

CHR: chromosome

POS: position (hg38)

MarkerID: the genetic variant being tested

Allele1: reference allele

Allele2: alternate allele

AC_Allele2: Alternate allele count

AF_Allele2: Alternate allele frequency in the sample

MissingRate: genotype missing rate

BETA: Effect size estimate (per allele effect)

SE: Standard error of beta

Tstat: T-statistic for the effect size

var: Variance of the score test statistic

p.value: P-value

p.value.NA: P-value computed using a normal approximation (Wald or score test), not SPA.

Is.SPA:  Whether the Saddlepoint Approximation (SPA) converged (TRUE/FALSE)

N: Sample size used in the test

p_bonf: P-value adjusted for the number of variants tested in that feature (Bonferroni)

p_FDR: P-value adjusted for the number of variants tested in that feature (FDR)

ACAT_P: Aggregated Cauchy Association Test (ACAT) p-value, which is a gene-level p-value combining multiple single-variant p-values for a given gene or region.