Table 2 Description of required and optional arguments for eQTpLot

Argument

Description

eQTL.df

A data frame of eQTL summary statistic data, as defined in Table 1

GWAS.df

A data frame of GWAS summary statistic data, as defined in Table 1

gbuild

Default value is “hg19”. The genome build, in quotes, to use for fetching genomic information for the genome track (panel B). This build should match the genome build used for “CHR” and “BP” in the GWAS.df. Currently the only compatible options are “hg19” and “hg38”

gene

The name/symbol of the gene to analyze, in quotes (Note: gene name must match an entry in Genes.df for the specified gbuild)

sigpvalue_eQTL

Default value is 0.05. The significance threshold to use for eQTL data (variants with an eQTL p-value larger than this threshold will be excluded from the analysis)

sigpvalue_GWAS

Default value is 5e-8. The significance threshold to use for GWAS data (this value will be used for a horizontal line in plot A, and to define GWAS significant/non-significant variants for the eQTL enrichment plot).

tissue

Default value is “all”. The tissue name, in quotes, to use for analysis. eQTL.df entries will be filtered to contain only data on this tissue. If this parameter is set to “all”, eQTpLot will pick the smallest eQTL p-value for each SNP across all tissues for a PanTissue analysis. Alternatively, a list of tissue names can be supplied (in the format c(“tissue1”, “tissue2”, …) to perform a PanTissue analysis on just these tissues. (Note: the tissue name must match at least one entry in the eQTL.df Tissue column)

trait

The name of the GWAS phenotype to analyze, in quotes. If all the data in GWAS.df is for a single phenotype and no PHE column is present, this argument will be used as the name for the analyzed phenotype. If GWAS.df contains information on multiple phenotypes, as specified in the optional GWAS.df PHE column, this parameter will be used to filter in GWAS.df entries for only this phenotype.

Argument

Description

Genes.df

A data frame of gene coordinates, as defined in Table 1

LD.df

A data frame of pairwise linkage data, as defined in Table 1

congruence

Default value is FALSE. If set to TRUE, variants with congruent and incongruent effects will be plotted separately, as described below.

genometrackheight

Default value is 2 Used to set the height of the genome track panel (B). Gene-dense regions may require more plotting space, whereas gene-sparse regions may look better with less plotting space.

getplot

Default value is TRUE. If set to FALSE, eQTpLot will not display the generated plot in the viewport.

LDcolor

Only used if LD.df is supplied. Default value is “color”. For the LDheatmap panel, the heatmap will be filled using a grayscale palate if this argument is set to “black”, or with a full color palate if this argument is set to “color”.

LDmin

Only used if LD.df is supplied. Default value is 10. For the LDheatmap panel, only variants that are in LD (with R² > R2min) with at least this many other variants will be displayed. This parameter can be useful to thin the number of variants being plotted in the LDheatmap.

leadSNP

Only used if LD.df is supplied. This parameter is used to specify the lead SNP ID, in quotes, to use for plotting LD information in the P-P plots. The specified variant must be present in both the GWAS.df and LD.df data frames.

NESeQTLRange

the maximum and minimum limits in the format c (min,max), to display for the NES value in eQTL.df. The default setting will adjust the size scale automatically to fit the displayed data, whereas specifying the limits will keep them consistent between plots.

R2min

Only used if LD.df is supplied. Default value is 0.1. The threshold for R² to use when selecting LD data from LD.df. Variant pairs with R² < R2min will not be included in the analysis.

range

Default value is 200. The range, in kB, to extend the analysis window on either side of the gene of interest, as defined by the Start and Stop points for the specified gene in Genes.df.

res

Default value is 300. The resolution, in dpi, for the output plot image

saveplot

Default value is TRUE. If set to TRUE, eQTpLot will save the generated plot in the working directory with the name “gene.trait.tissue. Congreunce_Info.LD_Info.eQTpLot.png”, using the variables and arguments provided.

wi

Default value is 12 if LD.df is not supplied, 14 if LD.df is supplied. The width of the output plot image, in inches. The height of the plot is calculated from this argument as well to maintain the appropriate aspect ratio.

xlimd

used to manually adjust the x axis maximum for the P-P plot, if needed

ylima

used to manually adjust the y axis maximum in plot A, if needed

ylimd

used to manually adjust the y axis maximum for the P-P plot, if needed

CollapsMethod

Default value is “min”. This parameter dictates the method used to collapse eQTL p-values and NES across tissues if a MultiTissue or PanTissue analysis is specified. If set to “min” the p-value and NES from the tissue with the smallest p-value for each variant will be selected. If set to “median” or “mean” the median or mean p-value and NES for each variant, across all specified tissues, will be selected. If set to “meta” eQTpLot will perform a simple sample-size-weighted meta-analysis [22, 23] of the p-values across all specified tissues.

(NOTE: If “meta” is specified, eQTL.df should include a column with header “N” indicating the number of samples used to derive the given eQTL data. If no column N is present, eQTpLot will give the user the option to complete a meta-analysis assuming equal sample sizes for all tissues, which may lead to inaccurate results. Also note that if “meta” is specified, no meta-analyzed NES will be computed, and all variants will be displayed as the same size in the main eQTpLot figure.)

Gene.List

Default value is FALSE. If set to TRUE, this parameter will output the Pearson correlation between eQTL and GWAS p-values for a given tissue across a user-supplied list of genes, ordered by significance. No plots will be generated. If the user sets the parameter tissue to “all,” or to a list of tissues, eQTpLot will collapse the eQTL data for these tissues by variant, using the method specified by the parameter CollapseMethod. This may be a useful parameter to obtain a very simple bird’s-eye view of the genes at a locus whose expression is most closely correlated to a relevant GWAS signal for a given trait.

Tissue.List

Default value is FALSE. If set to TRUE, this parameter will output the Pearson correlation between eQTL and GWAS p-values for a given gene across a user-supplied list of tissues, ordered by significance. No plots will be generated. If the user sets the parameter tissue to “all,” eQTpLot will consider each tissue included in eQTL.df. This may be a useful parameter to obtain a very simple bird’s-eye view of the tissues in which a given gene’s expression is most closely tied to a relevant GWAS signal for a given trait.