Users can specify a prefix for FastENLOC output files using the -prefix output_prefix option. If this option is not provided, FastENLOC will output files to the current directory with default names:
The specified prefix can include directory information. For example:
-prefix output/test
This will save all output files in the output subdirectory, each beginning with the prefix test. Specifying an output prefix is strongly recommended, as default output files may be overwritten by subsequent FastENLOC runs.
The enrichment parameter estimates are saved in a text file with the suffix enloc.enrich.out. An example of such a file is shown below:
Intercept -7.663 -
Enrichment (no shrinkage) 4.421 0.147
Enrichment (w/ shrinkage) 4.242 0.144
## Alternative (coloc) parameterization: p1 = 4.693e-04, p2 = 7.598e-04, p12 = 2.483e-05
where:
Intercept: the point estimate of $\alpha_0$Enrichment: the point estimate of $\alpha_1$, followed by the corresponding standard errorBoth the non-shrinkage and shrinkage versions of the $\alpha_1$ estimate are included; however, the shrinkage estimate is used for calculating colocalization probabilities.
The last line provides the equivalent transformation to the $(p_1, p_2, p_{12})$ priors used in coloc.
The text tabular file with the suffix enloc.mi.out contains the parameter estimates from each multi-imputation run.
A header is included.
The top lines from an example output are shown below:
a0 a1 p_eqtl p_gwas
-7.646 4.416 7.696e-04 5.092e-04
-7.648 4.347 7.919e-04 5.069e-04
-7.648 4.332 7.813e-04 5.061e-04
-7.646 4.354 7.691e-04 5.074e-04
-7.644 4.296 7.813e-04 5.074e-04
The colocalization probability for each variant within a signal cluster, credible set, or locus is saved in a tab-delimited text file with the suffix enloc.snp.out. A header row is included in the file. Below are the top lines from an example file:
Signal SNP PIP_qtl PIP_gwas_marginal PIP_gwas_qtl_prior SCP
ENSG00000006837:1(@)ENSG00000006837 ENSG00000006837_chr5_133528592_C_A_b38 4.763e-02 3.730e-05 1.491e-04 1.158e-04
ENSG00000006837:1(@)ENSG00000006837 ENSG00000006837_chr5_133528604_C_T_b38 1.391e-01 3.830e-05 3.781e-04 3.472e-04
ENSG00000006837:1(@)ENSG00000006837 ENSG00000006837_chr5_133528881_G_A_b38 1.501e-01 3.757e-05 3.974e-04 3.675e-04
ENSG00000006837:1(@)ENSG00000006837 ENSG00000006837_chr5_133529047_C_T_b38 1.501e-01 3.757e-05 3.974e-04 3.675e-04
ENSG00000006837:1(@)ENSG00000006837 ENSG00000006837_chr5_133529522_C_T_b38 8.186e-02 3.778e-05 2.340e-04 2.015e-04
The columns are:
Signal: ID of the corresponding signal cluster, credible set, or locusSNP: variant identifierPIP_qtl: the posterior probability of the variant being a QTL, denoted as $P(\gamma = 1 \mid {\rm QTL~ data})$PIP_gwas_marginal: the posterior probability of the variant being a GWAS hit, denoted as $P(d = 1 \mid {\rm GWAS~ data})$PIP_gwas_qtl_prior: the probability of the variant being a GWAS hit, conditional on it being a QTL, denoted as $P(d = 1 \mid {\rm GWAS~ data}, \gamma=1)$SCP: SNP-level colocalization probability, denoted as $P(d=1, \gamma=1 \mid {\rm GWAS~ data,~ QTL ~ data})$The colocalization probability for each signal cluster, credible set, or locus is saved in a tab-delimited text file with the suffix enloc.sig.out. A header row is included in the file. Below are the top lines from an example file:
Signal Num_SNP CPIP_qtl CPIP_gwas_marginal CPIP_gwas_qtl_prior RCP LCP
ENSG00000006837:1(@)ENSG00000006837 18 9.894e-01 6.649e-04 3.003e-03 2.472e-03 3.053e-03
ENSG00000006837:2(@)ENSG00000006837 1 2.101e-02 1.487e-04 3.346e-04 1.978e-04 1.978e-04
ENSG00000011083:1(@)ENSG00000011083 1 9.953e-01 4.443e-05 2.881e-03 2.881e-03 2.881e-03
ENSG00000011083:2(@)ENSG00000011083 10 9.953e-01 1.865e-03 1.076e-02 9.157e-03 1.075e-02
ENSG00000011083:3(@)ENSG00000011083 4 1.531e-02 7.739e-04 9.195e-04 2.303e-04 2.383e-04
The columns are:
Signal: ID of the signal cluster, constructed by concatenating the signal cluster IDs from the QTL and GWAS annotations, separated by (@)Num_SNP: the number of the member SNPs in the signal clusterPIP_qtl: the cumulative posterior probability that the signal cluster contains a QTL, denoted as $\sum_i P(\gamma_i = 1 \mid {\rm QTL~ data})$, where variant $i$ is a member SNPCPIP_gwas_marginal: the cumulative posterior probability that the signal cluster contains a causal GWAS hit, denoted as $\sum_i P(d_i = 1 \mid {\rm GWAS~ data})$, where variant $i$ is a member SNPCPIP_gwas_qtl_prior: the cumulative posterior probability that the signal cluster contains a causal GWAS hit, conditional on the hit also being a QTL, denoted as
$\sum_i P(d_i = 1 \mid {\rm GWAS~ data}, \gamma_i=1)$, where variant $i$ is a member SNPRCP: Signal/Regional-level colocalization probability, denoted as $\sum_i P(d_i=1, \gamma_i=1 \mid {\rm GWAS~ data,~ QTL~ data})$, where variant $i$ is a member SNPLCP: Locus-level colcalization probability, denoted as $\sum_{i,j} P(d_i=1, \gamma_j=1 \mid {\rm GWAS~ data,~ QTL~ data})$, where variants $i$ and $j$ are member SNPs.Note that LCP is always no smaller than RCP.
If the signal clusters in the QTL annotation follows the naming convention gene_name:signal_id, FastENLOC will generate a gene-level colocalization probability output accumulating signal-level colocalization probabilities across the gene with the ID gene_name.
This file has a suffix enloc.gene.out and a header row is included.
Below are the top lines from an example file:
Gene GRCP GLCP
ENSG00000006837 2.669e-03 3.250e-03
ENSG00000011083 1.224e-02 1.384e-02
ENSG00000013561 3.228e-04 3.228e-04
ENSG00000015479 1.092e-03 1.094e-03
ENSG00000016082 5.946e-03 7.906e-03
ENSG00000019582 1.191e-03 1.191e-03
The columns are:
Gene: gene IDGRCP: the probability that the gene contains at least one signal cluster with RCP $ \gt 0$GLCP: the probability that the gene contains at least one signal cluster with LCP $ \gt 0$Note that GRCP value quantifies the probability that the gene contains at least one colocalized SNP.