xMAGIC is a scalable method designed to integrate a large number of multi-omic xQTL datasets across diverse biological contexts with genome-wide association studies (GWAS) summary statistics. By linking epigenetic marks to target genes using multiple complementary approaches (e.g., chromatin interaction maps, pleiotropy association analysis) and combining association signals from expression and epigenetic phenotypes into a unified gene-trait association test, xMAGIC facilitates the identification of putative effector genes for 75.4% of GWAS loci, as demonstrated in an analysis of 45 human complex traits and 428 xQTL datasets, providing mechanistic insights into genetic associations.
You can download the executable binary in release page.
wget https://github.com/jianyanglab/MAGIC/releases/download/v1.0.2/xmagic-linux
chmod 754 xmagic-linux
Verify the installation by running the following command:
xmagic-linux --help
xMAGIC integrates GWAS and xQTL summary statistics in a single-step analysis.
We have curated and prepared a variety of publicly available molecular QTL data (downloadable from the yanglab website) and functional element-to-gene maps (downloadable from the yanglab website), which users can use to perform the xMAGIC analysis with their specific complex trait of interest. For illustration purposes, we provide demonstration data that can be used to run xMAGIC analysis with the command line below.
To get started quickly, you can download example data files:
# Download the complete example dataset
mkdir example_data
cd example_data
wget https://yanglab.westlake.edu.cn/data/magic-portal/example_data/GWAS_data.tar.gz
wget https://yanglab.westlake.edu.cn/data/magic-portal/example_data/e2g_data.tar.gz
wget https://yanglab.westlake.edu.cn/data/magic-portal/example_data/LD_reference.tar.gz
wget https://yanglab.westlake.edu.cn/data/magic-portal/example_data/xQTL_data.tar.gz
for f in *.tar.gz; do tar -xvzf "$f"; doneThe basic command to run xMAGIC is:
./xmagic-linux --besd-flist example_data/xQTL_data/user_xQTL_list.txt \
--gwas-summary example_data/GWAS_data/GWAS_T2D_chr11.txt \
--bfile example_data/LD_reference/ALL_1KGP3_Phase3_mind95_geno95_maf01_hwe1e06_chr \
--e2g-flist example_data/e2g_data/user_e2g_list.txt \
--chr 11 \
--trait-name T2D \
--out example_data/myxmagic_test
A text file that lists the paths to multiple xQTL BESD datasets (same format as SMR: https://yanglab.westlake.edu.cn/software/smr/#DataManagement).
These xQTL datasets can span multiple omics layers,such as transcriptomics (eQTL), proteomics (pQTL), and epigenomics (mQTL), and can be generated from diverse cellular or environmental conditions.
Example
example_data/xQTL_data/user_xQTL_list.txt:
eQTL_DIRECT example_data/xQTL_data/eQTL_DIRECT_chr TRUE gene
eQTL_GTEx_Whole_Blood example_data/xQTL_data/eQTL_GTEx_Whole_Blood_chr TRUE gene
pQTL_FENLAND example_data/xQTL_data/pQTL_FENLAND_chr TRUE gene
mQTL_BrainMeta example_data/xQTL_data/mQTL_BrainMeta_chr TRUE epigenetic
hQTL_H3K4me1_BLUEPRINT example_data/xQTL_data/hQTL_H3K4me1_BLUEPRINT_chr TRUE epigenetic
caQTL_Blood example_data/xQTL_data/caQTL_Blood_chr TRUE epigenetic
Columns are:
- QTL name — for expression QTLs, use the prefix
eQTL_*; for splicing QTLs, usesQTL_*, etc. - Path — path to the BESD-formatted QTL file. It is recommended to use absolute paths to avoid issues when running the program from different working directories.
- Chromosome split flag —
TRUEif the xQTL datasets are split by chromosome,FALSEotherwise. - QTL type — either
geneorepigenetic. For entries marked asepigenetic, a probe-to-gene link file is required so that signals can be mapped to target genes, please see the--e2g-flistflag.
GWAS summary statistics file (format similar to GCTA-COJO: https://yanglab.westlake.edu.cn/software/gcta/#COJO).
PLINK binary files (.bed, .bim, .fam) for LD reference.
A text file listing paths to multiple functional element-to-gene mapping files. Each mapping file should be tab-delimited with at least four columns: the first three for the element coordinates (chromosome, start position, end position) and the fourth for the gene name. An optional fifth column can specify association strength, and an optional sixth column can specify context. Example file list (mye2g.list):
Example
example_data/e2g_data/user_e2g_list.txt
ABC example_data/e2g_data/ABC_chr11.bed
EpiMap example_data/e2g_data/EpiMap_chr11.bed
RoadMap example_data/e2g_data/RoadMap_chr11.bed
PCHiC example_data/e2g_data/PCHiC_chr11.bed
Promoter example_data/e2g_data/Promoter_chr11.bed
closestTSS example_data/e2g_data/closestTSS_chr11.bed
Columns are:
- epigenetic link name — the enhancer-promoter interaction resource name
ABC, etc. - Path — path to the enhancer-promoter interaction file. It is recommended to use absolute paths to avoid issues when running the program from different working directories.
Example mapping file example_data/e2g_data/ABC_chr11.bed:
chr11 100438339 100438739 PGR 0.0258146262233206 A Cardiomyocyte
chr11 100438339 100438739 PGR 0.0271777153527535 A Cardiomyocyte
chr11 100439170 100439570 PGR 0.0192453725099139 A Cardiomyocyte
chr11 100439170 100439570 PGR 0.0202615854828699 A Cardiomyocyte
chr11 100445172 100445572 CNTN5 0.0335977399341998 A Cardiomyocyte
chr11 100445172 100445572 CNTN5 0.0340138578504139 A Cardiomyocyte
chr11 100472725 100473125 BIRC2 0.0188125575399639 A Cardiomyocyte
chr11 100472725 100473125 BIRC2 0.0185803409452074 A Cardiomyocyte
chr11 100576412 100576812 CNTN5 0.0194293770614472 A Cardiomyocyte
chr11 100576412 100576812 CNTN5 0.0196700156255881 A Cardiomyocyte
If you don’t have an epigenetic–gene link file, we provide a set of mapping methods, including ABC, Roadmap, EpiMap, PCHiC, Promoter, and Closest TSS that were used in our manuscript.
wget https://yanglab.westlake.edu.cn/data/magic-portal/example_data/epimark_to_gene_links.tar.gz
tar -xvzf epimark_to_gene_links.tar.gz
Prefix for output files (example_data/myxmagic_test), including gene-trait association p-values.
Example Output example_data/myxmagic_test_20251105_155430/MAGIC/summary/T2D_MAGIC.txt:
chr start end strand gene_id gene_name GWAS_LOCUS Lead_SNP Lead_SNP_BP MAGIC eMAGIC sMAGIC pMAGIC edMAGIC mMAGIC hMAGIC caMAGIC eMAGIC_QTL_name sMAGIC_QTL_name pMAGIC_QTL_name edMAGIC_QTL_name mMAGIC_QTL_name hMAGIC_QTL_name caMAGIC_QTL_name eMAGIC_probeID sMAGIC_probeID pMAGIC_probeID edMAGIC_probeID mMAGIC_probeID hMAGIC_probeID caMAGIC_probeID
11 16613132 16758340 + ENSG00000110696 C11orf58 chr11:16388025:18388025 rs5219 17388025 0.00512866875067353 0.00763007269639537 NA NA NA 0.00386236898797709 NA NA eQTL_DIRECT NA NA NA mQTL_BrainMeta NA NA ENSG00000110696 NA NA NA cg17284609 NA NA
11 16777297 17014414 - ENSG00000166689 PLEKHA7 chr11:16388025:18388025 rs5219 17388025 7.55413644832892e-05 0.000149159376880648 NA 0.00554274494147966 NA 1.78113700000093e-05 0.9214682 0.000315176199999989 eQTL_DIRECT NA pQTL_FENLAND NA mQTL_BrainMeta hQTL_H3K4me1_BLUEPRINT caQTL_Blood ENSG00000166689 NA X12731_12_PLEKHA7 NA cg02095290 11:16932882:16934884 197402
11 17074388 17077715 - ENSG00000110700 RPS13 chr11:16388025:18388025 rs5219 17388025 0.058802868840196 0.058802868840196 NA NA NA NA NA NA eQTL_GTEx_Whole_Blood NA NA NA NA NA NA ENSG00000110700.6 NA NA NA NA NA NA
11 17077730 17207986 - ENSG00000011405 PIK3C2A chr11:16388025:18388025 rs5219 17388025 1.44278256009045e-10 1.44278256009045e-10 NA NA NA NA NA NA eQTL_DIRECT NA NA NA NA NA NA ENSG00000011405 NA NA NA NA NA NA
11 17208153 17349980 + ENSG00000070081 NUCB2 chr11:16388025:18388025 rs5219 17388025 1.60460533749074e-12 1.60460533749074e-12 NA NA NA NA NA NA eQTL_DIRECT NA NA NA NA NA NA ENSG00000070081 NA NA NA NA NA NA
11 17351800 17377341 + ENSG00000188211 NCR3LG1 chr11:16388025:18388025 rs5219 17388025 9.42890210353653e-12 0.1771537 NA NA NA 4.71445105176826e-12 NA NA eQTL_DIRECT NA NA NA mQTL_BrainMeta NA NA ENSG00000188211 NA NA NA cg16820186 NA NA
11 17365172 17389331 - ENSG00000187486 KCNJ11 chr11:16388025:18388025 rs5219 17388025 3.01144472537199e-24 4.04787314778332e-13 NA NA NA 1.5057223626916e-24 NA NA eQTL_DIRECT NA NA NA mQTL_BrainMeta NA NA ENSG00000187486 NA NA NA cg15432903 NA NA
11 17392498 17476894 - ENSG00000006071 ABCC8 chr11:16388025:18388025 rs5219 17388025 0.967843892880911 NA NA NA NA 0.967843892880911 NA NA NA NA NA NA mQTL_BrainMeta NA NA NA NA NA NA cg26105227 NA NA
11 17493895 17544416 - ENSG00000006611 USH1C chr11:16388025:18388025 rs5219 17388025 0.464564995136756 NA NA NA NA 0.464564995136756 NA NA NA NA NA NA mQTL_BrainMeta NA NA NA NA NA NA cg14425863 NA NA
11 17547259 17647150 + ENSG00000188162 OTOG chr11:16388025:18388025 rs5219 17388025 0.872053797891558 NA NA NA NA 0.872053797891558 NA NA NA NA NA NA mQTL_BrainMeta NA NA NA NA NA NA cg17099016 NA NA
11 17719571 17722136 + ENSG00000129152 MYOD1 chr11:16388025:18388025 rs5219 17388025 0.03065185 NA NA NA NA 0.03065185 NA NA NA NA NA NA mQTL_BrainMeta NA NA NA NA NA NA cg01144436 NA NA
11 17734774 17856804 + ENSG00000129159 KCNC1 chr11:16388025:18388025 rs5219 17388025 0.0194788299999999 NA NA NA NA 0.0194788299999999 NA NA NA NA NA NA mQTL_BrainMeta NA NA NA NA NA NA cg19244655 NA NA
11 17788048 18013047 - ENSG00000129158 SERGEF chr11:16388025:18388025 rs5219 17388025 0.0792508312751415 0.0492783969930815 NA NA NA 0.0937100832877356 NA 0.1418364 eQTL_DIRECT NA NA NA mQTL_BrainMeta NA caQTL_Blood ENSG00000129158 NA NA NA cg23625312 NA 197481
11 18017555 18046269 - ENSG00000129167 TPH1 chr11:16388025:18388025 rs5219 17388025 0.0435363618627685 0.0435363618627685 NA NA NA NA NA NA eQTL_DIRECT NA NA NA NA NA NA ENSG00000129167 NA NA NA NA NA NA
11 18069935 18106087 - ENSG00000166788 SAAL1 chr11:16388025:18388025 rs5219 17388025 0.0247461308158468 0.0243330464849628 NA NA NA 0.02561576 NA NA eQTL_DIRECT NA NA NA mQTL_BrainMeta NA NA ENSG00000166788 NA NA NA cg12377368 NA NA
11 18120955 18138488 + ENSG00000179826 MRGPRX3 chr11:16388025:18388025 rs5219 17388025 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA
11 18172837 18174280 + ENSG00000179817 MRGPRX4 chr11:16388025:18388025 rs5219 17388025 0.177807000911694 NA NA NA NA 0.177807000911694 NA NA NA NA NA NA mQTL_BrainMeta NA NA NA NA NA NA cg02608124 NA NA
11 18231355 18236802 - ENSG00000148965 SAA4 chr11:16388025:18388025 rs5219 17388025 0.462144826207657 NA NA 0.462144826207657 NA NA NA NA NA NA pQTL_FENLAND NA NA NA NA NA NA X15516_12_SAA4 NA NA NA NA
11 18231423 18248635 - ENSG00000255071 SAA2-SAA4 chr11:16388025:18388025 rs5219 17388025 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA
11 18239223 18248668 - ENSG00000134339 SAA2 chr11:16388025:18388025 rs5219 17388025 0.0571837028787258 NA NA 0.0571837028787258 NA NA NA NA NA NA pQTL_FENLAND NA NA NA NA NA NA X18832_65_SAA2 NA NA NA NA
11 18266260 18269977 + ENSG00000173432 SAA1 chr11:16388025:18388025 rs5219 17388025 0.110115525667891 NA NA 0.0545737222970077 NA 0.5709017 NA NA NA NA pQTL_FENLAND NA mQTL_BrainMeta NA NA NA NA X15515_2_SAA1 NA cg24015472 NA NA
11 18278668 18322198 - ENSG00000110756 HPS5 chr11:16388025:18388025 rs5219 17388025 0.0422314189381126 0.0422314189381126 NA NA NA NA NA NA eQTL_GTEx_Whole_Blood NA NA NA NA NA NA ENSG00000110756.17 NA NA NA NA NA NA
11 18322295 18367045 + ENSG00000110768 GTF2H1 chr11:16388025:18388025 rs5219 17388025 0.407604333637021 0.407604333637021 NA NA NA NA NA NA eQTL_GTEx_Whole_Blood NA NA NA NA NA NA ENSG00000110768.11 NA NA NA NA NA NA
Columns include chromosome, start/end positions, strand, gene ID, gene symbol, GWAS locus region, overall xMAGIC p-value, and layer-specific p-values for eMAGIC, sMAGIC, pMAGIC, edMAGIC, mMAGIC, hMAIGIC, caMAGIC (eQTL, sQTL, pQTL, edQTL, caQTL, mQTL, hQTL), with NA indicating unavailable data, as well as the most significant QTL name and probeID for each omics layer.
xMAGIC shares data formats with SMR. For full options, see SMR documentation. Curated xQTL datasets (428 total) and the functional element-to-gene maps can be downloaded from the yanglab website.
An online platform for xMAGIC is available at xMAGIC Portal (http://yanglab.westlake.edu.cn/xMAGIC), which requires only the upload of GWAS summary statistics (try example_data/GWAS_data/GWAS_T2D_chr11.txt).
Qi T, Guo Y, Chen C, Xu T, Luo J, Jiang Z, Chen H, Guo M, Wang K, Hou J, Yang J. (2025). Integrative analysis of hundreds of multi-omic and cross-context xQTL datasets links over three-quarters of GWAS loci to putative effector genes. Under Review.
Ting Qi ([email protected]) or Jian Yang ([email protected]).