Autoreporting results format

utoreporting outputs two types of tab-separated results: variant and group reports. Group reports contain information about the groups built around credible sets, with one credible set per row. They provide information about the credible set and its lead variant combined with various annotations.

The variant reports list all of the variants in the group reports' groups. These variants include credible set lead variants, as well as variants that were LD clumped together with the group lead variant. These are also combined with a set of annotations. For more information about the Autoreporting tool and how it works, see Autoreporting in FinnGen.

Columns can vary between releases as the autoreporting pipeline has evolved to include more information while other data has been removed due to becoming redundant. The formats described below are for FinnGen release 13 (R13).

Variant reports

The variant report is a tab-delimited (.tsv) file summarizing the results of the endpoint’s autoreporting run. It details association statistics for all lead variants and their LD-clumped partners across every credible set. Annotations are sourced from results from the previous FinnGen release, the gnomAD database and the GWAS Catalog. The columns are as follows:

Column	Description	Example value/Formatting
#chrom	chromosome of variant	1
pos	variant position	123987
ref	variant reference allele	A
alt	variant alternate allele	C
pval	variant p-value	5.01e-7
beta	effect size of alternate allele	0.025026
r2_to_lead	LD r2 value between this variant and the credible set's lead variant	0.85479
mlogp	-log10 p-value of variant	6.30016
sebeta	standard error of effect size	0.004979
af_alt	allele frequency of the alternate allele	0.405409
af_alt_cases	allele frequency of the alternate allele in cases	0.411388
af_alt_controls	allele frequency of the alternate allele in controls	0.403217
cs_id	credible set id	chr1_123456_A_C_1
cs_region	finemapping region	1:1-30000001
cs_number	credible set number in finemapping region	4
cs_prob	variant's posterior inclusion probability (PIP) representing the likelihood of it being the true causal variant in it's credible set	0.0629503
cs_log10bf	log10 Bayes factor for finemapping solution including this credible set vs. solution not including it	3.96812
cs_min_r2	minimum LD r2 between variants in the credible set	0.46702
cs_size	credible set size	5
good_cs	Whether the credible set of this locus is of good quality. Currently good cs is one that has minimum LD value between credible set variants larger than 0.25.	True/False
#variant	variant ID	chr1_123987_A_C
locus_id	The locus in question, formatted from the top SNP's chromosome, position, reference and alternate alleles. In case of credible set grouping, the top SNP is the variant with the largest PIP in that credible set. In case of LD and simple grouping, the top SNP is the variant with smallest p-value of that group/region. Most if not all release results are grouped around credible sets.	chr1_1_C_T for a lead variant with chromosome 1, position 1, reference allele C and alternate allele T.
pos_rmax	maximum basepair position of locus, including credible set and LD partners	30000001
pos_rmin	minimum basepair position of locus, including credible set and LD partners	1
phenotype	phenotype code	-
longname	phenotype description	-
category	phenotype's ICD chapter or FinnGen endpoint category	II Neoplasms, from cancer register (ICD-O-3)
n_cases	number of cases for this phenotype	1234
n_controls	number of controls for this phenotype	1234
enrichment_nfsee	same as GENOME_FI_enrichment_nfe_est column	0.888291
fin.AF	variant Finnish-ancestry allele frequency from gnomAD genome-seq data	0.123
fin.AN	called allele count (either ref or alt) for variant in Finnish-ancestry samples, sourced from gnomAD genome-seq data	1234
fin.AC	variant Finnish-ancestry allele count from gnomAD genome-seq data	123
fin.homozygote_count	number of homozygote carriers in Finnish-ancestry samples from gnomAD genome-seq data	12
fet_nfsee.odds_ratio	odds ratio of Fischer's exact test for enrichment of alt allele in Finnish-ancestry vs. non-Finnish, Swedish or Estonian Europeans, based on gnomAD genome-seq data	1.15
fet_nfsee.p_value	p-value of Fischer's exact test for enrichment of alt allele in Finnish-ancestry vs. non-Finnish, Swedish or Estonian Europeans, based on gnomAD genome-seq data	5.01e-3
nfsee.AC	variant non-Finnish, Swedish or Estonian European allele count from gnomAD genome-seq data	123
nfsee.AN	called allele count (either ref or alt) for variant in non-Finnish, Swedish or Estonian Europeans, sourced from gnomAD genome-seq data	1234
nfsee.AF	variant non-Finnish, Swedish or Estonian Europeans allele frequency from gnomAD genome-seq data	0.123
nfsee.homozygote_count	number of homozygote carriers in non-Finnish, Swedish or Estonian Europeans from gnomAD genome-seq data	123
most_severe_gene*	most severe gene of the variant	APOE
most_severe_consequence	most severe consequence of variant	missense_variant
FG_INFO	variant imputation INFO score in this FinnGen release	0.9951358663750908
n_INFO_gt_0_6
functional_category	variant functional category from gnomAD genome-seq data	pLoF
rsids	rsids of this variant	rs1234
GNOMAD_AF_fin	Finnish allele frequency from gnomAD 4.1	0.45
GNOMAD_AF_nfe	non-Finnish European allele frequency in gnomAD 4.1	0.45
GNOMAD_FI_enrichment_nfe	Finnish enrichment of the variant compared to non-Finnish European population in gnomAD 4.1	1.2
pval_previous_release	variant p-value for same endpoint in previous FinnGen release	3.55901e-11
beta_previous_release	variant effect size of alternate allele for same endpoint in previous FinnGen release	0.042283
#variant_hit	same as #variant column if variant has been previously reported as a GWAS hit	chr1_123987_A_C
pval_trait	p-value of variant for previously reported GWAS hit	6e-32
trait	EFO ID of trait for which variant has been previously reported as a GWAS hit	EFO_0005711
trait_name	name/description of trait for which variant has been previously reported as a GWAS hit	household income
study_link	pubmed link to study where variant has been previously reported as a GWAS hit (if available)	www.ncbi.nlm.nih.gov/pubmed/01234567

Group reports

The group report is a tab-delimited (.tsv) file and contains the group-level (credible set) summary of an autoreporting run. They are aggregated from the variant reports, with one credible set per row. Annotations are sourced from results from the previous FinnGen release, the gnomAD database and the GWAS Catalog. The columns are as follows:

Column	Description	Example value/Formatting
phenotype	phenotype name	-
phenotype_abbreviation	phenotype code	-
locus_id	The locus in question, formatted from the top SNP's chromosome, position, reference and alternate alleles. In case of credible set grouping, the top SNP is the variant with the largest PIP in that credible set. In case of LD and simple grouping, the top SNP is the variant with smallest p-value of that group/region. Most if not all release results are grouped around credible sets.	chr1_1_C_T for a lead variant with chromosome 1, position 1, reference allele C and alternate allele T.
rsids	rsids of this variant	rs1234
Cases	Number of cases for this phenotype	1234
Controls	Number of controls for this phenotype	1234
chrom	chromosome of locus	1
pos	lead variant position	123456
ref	lead variant reference allele	A
alt	lead variant alternate allele	C
pval	lead variant p-value	5.01e-7
lead_r2_threshold	minimum LD r2 with lead variant for inclusion in this locus, calculated based on a minimum expected $\chi^2$ statistic of 5 - see lower half of this page for more detail.	0.13822
lead_beta_previous_release	effect size of alternate allele for same endpoint in previous FinnGen release	0.042283
lead_pval_previous_release	p-value for same endpoint in previous FinnGen release	3.55901e-11
lead_most_severe_consequence	most severe consequence of lead variant	missense_variant
lead_most_severe_gene*	most severe gene of the lead variant	APOE
lead_enrichment	How much the lead variant is enriched in Finnish population compared to Europe	4.35
lead_$COLUMN_NAME	other columns that are grabbed for the lead variant, such as effect size, standard error, p-value and allele frequencies	-
gnomAD_functional_category	functional category for the variant Exome data.	pLoF
gnomAD_enrichment_nfsee	lead variant enrichment in Finland against NFSEE (Europeans that are not Finnish, Swedish or Estonian) population from Exome data.	5.1
gnomAD_fin.AF	lead variant allele frequency in Finland. Exome data.	0.123
gnomAD_fin.AN	lead variant allele number in Finland. Exome data.	123
gnomAD_fin.AC	lead variant allele count in Finland. Exome data.	123
gnomAD_fin.homozygote_count	Amount of homozygote carriers in Finnish population. Exome data.	12
gnomAD_fet_nfsee.odds_ratio	Fischer's exact test for enrichment FIN vs. NFSEE odds ratio. Exome data.	1.15
gnomAD_fet_nfsee.p_value	Fischer's exact test for enrichment FIN vs. NFSEE p-value. Exome data.	5.01e-3
gnomAD_nfsee.AC	lead variant NFSEE population allele count. Exome data.	123
gnomAD_nfsee.AN	lead variant NFSEE population allele number. Exome data.	123
gnomAD_nfsee.AF	lead variant NFSEE population allele frequency. Exome data.	0.123
gnomAD_nfsee.homozygote_count	Amount of homozygote carriers in NFSEE population. Exome data.	123
cs_id	credible set id	chr1_123456_A_C_1
cs_size	credible set size	5
cs_log_bayes_factor	credible set bayes factor, log10	5.21
cs_number	credible set number in its region	1
cs_region	finemapping region	1:1-30000001
good_cs	Whether the credible set of this locus is of good quality. Currently good cs is one that has minimum LD value between credible set variants larger than 0.25.	True/False
credible_set_min_r2_value	Minimum LD r2 value between credible set variants	0.4
best_coding_var	The variant in credible set that has a functional consequence, and has the largest PIP.	chr1_1_A_T
best_coding_var_consequence	functional consequence for the best coding variant	missense_variant
best_coding_var_gene	gene in which the best coding variant has the consequence	GENE1
best_coding_var_af	Finnish allele frequency of the best coding variant	0.30685
best_coding_var_eur_af	non-finnish allele frequency of the best coding variant. Taken from gnomAD annotation.	0.32667
best_coding_var_beta	effect size of the best coding variant	0.059923
best_coding_var_p	p-value of best coding variant	4.16296e-5
start	locus start position in basepairs	1 for a group with positions [1,2,3,4,5]
end	locus end position in basepairs	5 for a group with positions [1,2,3,4,5]
found_associations_strict	This column lists all of the trait associations found in GWAS Catalog for variants that are in the credible set/strict group (strict group here means that in case of LD grouping, variants that are in higher LD than a given threshold, and have p-values lower than the significance threshold). The trait name is followed by the LD r² that reported variant had with the lead variant. If there are multiple variants associated with that trait, the largest value is chosen.	trait1|0.86;trait2|0.45
found_associations_relaxed	This column lists all of the trait associations found in GWAS Catalog for variants in the group. The trait name is followed by the LD r² to lead value of the variant that had the association. If there are multiple variants associated with that trait, the largest value is chosen.	trait1|0.86;trait2|0.45
credible_set_variants	This column lists the credible set variants. The PIP and R² values are listed after the variant	chr1_1_A_T|0.25|0.999
functional_variants_strict	All of the variants with a functional consequence, with the functional consequence label, gene and R² to lead variant. The variants are part of the credible set/strict group.	chr1_1_A_T|missense_variant|GENE1|0.45
functional_variants_relaxed	All of the variants with a functional consequence, with the functional consequence label and R² to lead variant. The variants are part of the credible set/strict group.	chr1_1_A_T|missense_variant|GENE1|0.45
specific_efo_trait_associations_strict	If specific traits were given to the script(e.g. equivalent EFO codes to the phenotype in question), any trait associations correspoding to those traits are listed here. This column lists only associations where the variant is in the credible set/strict group.	trait1|0.86;trait2|0.45
specific_efo_trait_associations_relaxed	If specific traits were given to the script(e.g. equivalent EFO codes to the phenotype in question), any trait associations correspoding to those traits are listed here. This column lists associations to all variants in the group.	trait1|0.86;trait2|0.45
n_ld_partners_0_8	number of nearby variants with LD r2$\ge$0.8 with lead variant	5
n_ld_partners_0_6	number of nearby variants with LD r2$\ge$0.6 with lead variant	23

*The HGNC symbols in the autoreporting files are version 38, as well as the VEP cache. "The most_severe_gene" and "most_severe" columns in the autoreporting file come from the finngen annotation file. Analysis team will update the release documentation to include this information (including those versions) for future releases.

For more information, see the release documentation:

/library-green/finngen_R8_analysis_documentation/

Read more about Autoreporting in FinnGen, and see also FAQ Do the autoreports report the 95% or 99% credible sets