NOTE: Development has moved to https://github.com/molgenis/systemsgenetics/tree/master/Decon2

Authors: Raúl Aguirre-Gamboa and Niek de Klein

Introduction

Decon-eQTL is part of Decon2, a statistical framework for estimating cell counts using molecular profiling such as expression or methylation data from heterogeneous samples (Decon-cell) and consecutive deconvolution of expression quantitative trait loci (Decon-eQTL) into each cell subpopulation.

Requirements

Java version 1.8 or higher (Tested using 1.8.0_121)
Decon-eQTL.jar

For compiling

commons-cli v1.3.1
commons-csv v1.2
commons-io v2.4
commons-lang3 v3.4
commons-math3 v3.6
jsci-core
jsci-mathmlimpl
jsci-sci

minimal usage example

java -jar deconvolution -c <file containing cellcounts> \
                        -e <file containing expression data> \
                        -g <file containing genotypes>  \
                        -o <output directory> \
                        -sn <file with SNP and gene combination to test>

Options overview

-ad,--all_dosages                         Filter out QTLs where not all dosages are present in at least 1 sample
-c,--cellcount <file>                     Cellcount file name
-cc,--use_relative_cellcounts             Calculate ratio between cellcount and cellcount average, use that as cellcount in the model
-e,--expression <file>                    Expression file name
-f,--filter_samples                       If set, remove samples that are filtered out because of -m or -ad. By default p-values of these are set to 333.0
-g,--genotype <file>                      Genotype file name
-gc,--genotypeConfigurationType <arg>     Which genotype configuration type to use (either all or two)
-help                                     print this message
-m,--minimum_samples_per_genotype <int>   The minimum amount of samples need for each genotype of a QTL for the QTL to be included in the results
-no,--no_console                          Do not output logging info to the console
-o,--outfolder <path>                     Path to folder to write output to
-oe,--outputPredictedExpression           Write output file with predicted expression
-of,--outfile <file>                      Outfile name of deconvolution results (will be written in outfolder)
-r,--round_dosage                         Round the dosage to the closest int
-s,--output_significant_only              Only output results that are significant in at least one celltype.
-sg,--skip_genotypes                      Skip genotypes that are in the GeneSNP pair file but not in the genotype file.
-sn,--snpsToTest <file>                   Tab delimited file with first column gene name, second column SNP name. Need to match with names from genotype and expression files.
-t,--test_run                             Only run deconvolution for 100 QTLs for quick test run
-w,--whole_blood_qtl                      Add whole blood eQTL (pearson correlation genotypes and expression)