In USCbiostats/hJAM: Hierarchical Joint Analysis of Marginal Summary Statistics

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)

JAM: Joint Analysis of Marginal summary statistics

About

In this package, we provide three JAM-family methods that utilize GWAS summary statistics to perform post-GWAS analysis, including Mendelian Randomization, TWAS, and multi-population fine-mapping.

hJAM

hJAM (hJAM) is a hierarchical model which unifies the framework of Mendelian Randomization and Transcriptome-wide association studies. The hJAM-Egger (hJAM_egger) is a natural extension on hJAM that uses an intercept term to account for the pleiotropy effect of the SNPs on the outcome.

Additionally, we provide implementations to construct the weight matrix A in hJAM. JAM_A converts the marginal summary statistics into conditional ones by using the correlation matrix of the SNPs from a reference data. susieJAM_A selects SNPs for one intermediate using the marginal summary statistics from GWAS or other study summary data.

SHA-JAM

SHA-JAM (SHAJAM) is a scalable version of hJAM that handles high-dimensional intermediates. SHA-JAM performs model selection from highly correlated intermediates through SuSiE: Sum of Single Effect Model (SHAJAM) or elastic-net (EN.hJAM).

mJAM

mJAM is for multi-population fine-mapping using GWAS summary statistics and credible set construction. We provide two implementations of mJAM: one through SuSiE (mJAM_SuSiE) and another one through forward selection (mJAM_Forward). mJAM_Forward also provides the flexiblity of constructing credible sets for using user-defined index SNPs. A tutorial to get started with mJAM can be found here.

Citing this work

• hJAM: Jiang, L., Xu, S., Mancuso, N., Newcombe, P. J., & Conti, D. V. (2021). A Hierarchical Approach Using Marginal Summary Statistics for Multiple Intermediates in a Mendelian Randomization or Transcriptome Analysis. American journal of epidemiology, 190(6), 1148–1158. https://doi.org/10.1093/aje/kwaa287

• SHA-JAM: Jiang, L., Conti, D.V. SHA-JAM: A Scalable Hierarchical Approach to Joint Analysis for Marginal Summary Statistics with Omics Data. (In preparation).

• mJAM: Shen, J., Jiang, L., Wang, K., Wang, A., Chen, F., Newcombe, P.J., Haiman, C.A., & Conti, D.V. Fine-Mapping and Credible Set Construction using a Multi-population Joint Analysis of Marginal Summary Statistics from Genome-wide Association Studies. (Preprint available)

Quick start with hJAM package

You can install the published version of hJAM from CRAN with:

install.packages("hJAM")

Currently we are working on improving hJAM package and adding genome-wide implementation of mJAM. If you want to use the most updated version of this package, we recommend installing the development version from GitHub with:

if (!require("devtools")) { install.packages("devtools") } else {}
devtools::install_github("USCbiostats/hJAM")

Session info

sessionInfo()

USCbiostats/hJAM documentation built on Jan. 26, 2024, 5:27 p.m.