METASOFT

METASOFT is a free, open-source meta-analysis software tool for genome-wide association study analysis, designed to perform a range of basic and advanced meta-analytic methods in an efficient manner.

ForestPMPlot is a free, open-source a python-interfaced R package tool for analyzing the heterogeneous studies in meta-analysis by visualizing the effect size differences between studies. The resulting plot can facilitate the better understanding of heterogeneous genetic effects on the phenotype in different study conditions.

Overview

METASOFT provides the following methods:

Fixed Effects model (FE): Fixed effects model based on inverse-variance-weighted effect size.
Random Effects model (RE): Conventional random effects model based on inverse-variance-weighted effect size (very conservative).
Han and Eskin's Random Effects model (RE2): New random effects model optimized to detect associations under heterogeneity. (Han and Eskin, AJHG 2011)
Binary Effects model (BE): New random effects model optimized to detect associations when some studies have an effect and some studies do not. (Han and Eskin, PLoS Genetics 2012)

METASOFT provides the following estimates:

Summary effect size estimates

Beta and standard errors for both fixed effects model and random effects model.

Heterogeneity estimates

Cochran's Q statistic and its p-value, and I²

M-values

Posterior probability that an effect exists in each study. (Han and Eskin, PLoS Genetics 2012)

ForestPMPlot is a visualization tool. The resulting plot consists of the following components:

Forest Plot

For each study, it displays p-value, study name, log odds ratio and its standard error and summary statistic.

m-value

m-value is the posterior probability that the effect exists in each study. m-values have the following interpretations.

Small m-value (e.g. < 0.1): The study is predicted to not have an effect.
Large m-value (e.g. > 0.9): The study is predicted to have an effect.
Otherwise: It is ambiguous to make a prediction.

m-value is estimated utilizing cross-study information (Han and Eskin, 2012 Link).

PM Plot

PM-Plot visualizes the m-value of each study along with its p-value. The X-axis of the PM-Plot represents the m-value between 0 and 1 and the Y-axis represents the statistical significance, −log10(p-value). PM-Plot allows us to predict which study have an effect and which study does not have an effect. The dot size represents the study sample size.

Download

Metasoft.zip containing the following:
- Metasoft.jar (java archive package file)
- HanEskinPvalueTable.txt (tabulated p-value file required to perform RE2)
- example.txt (Example meta-analysis data consisting of 3 studies)
- README
- LICENSE
- plink2metasoft.py (PLINK format converter; PLINK .assoc files → METASOFT format)
- plink2metasoft_subroutine.R (a subroutine file used by format converter)

Optional Java source code files (only if you need)
- Metasoft_src.zip

ForestPMPlot.zip containing the following:
- pmplot.py (main python script)
- forestpmplot.R (R ploting engine)
- data/input.txt (a sample MetaSoft input file )
- data/output.txt (a sample MetaSoft output file )
- study.name.txt (a sample file containing a list of study names)
- study.order.txt (a sample file containing a list of display indices)
- run.sh (a sample running script for generating Apoa2.pdf)
- Apoa2.pdf (a sample resulting ForestPMPlot)
- README
- LICENSE

You can SUBSCRIBE to our mailing list to be notified for any major updates and bug fixes.

Version/bug info

METASOFT

v2.0.1 (2012-08-21) PLINK format converter is added.
v2.0.0 (2012-02-15) M-values are computed using exact calculation. User interface has been changed.
v1.0.0 (2011-08-25) First major release. Bug fix.
v0.3.1 (2011-07-07) Bug fix.
v0.3.0 (2011-07-06) Bug fix. Binary effects model and M-values implemented.
v0.2.2 (2011-05-19) Bug fix.
v0.2.1 (2011-05-16) Small changes in output format.
v0.2.0 (2011-05-10) Bug fix. Genomic control method implemented.
v0.1.0 (2011-04-18) Initial prototype version deployed

ForestPMPlot

v1.0.1 (2014-10-24) Bug fix.
v1.0.0 (2014-08-12) First major release.

Publication

If you use our software or refer to the new random effects model method, please cite

Buhm Han and Eleazar Eskin, “Random-Effects Model Aimed at Discovering Associations in Meta-Analysis of Genome-wide Association Studies”, The American Journal of Human Genetics (2011) 88, 586-598. LINK

If you use m-values or the binary effects model method, please cite

Buhm Han and Eleazar Eskin, “Interpreting Meta-Analysis of Genome-wide Association Studies”, In Press. PLoS Genetics (2012). (Manuscript available here)

If you want to refer the GxE interpretation, please cite

Eun Yong Kang and Buhm Han and Nicholas Furlotte and Jong Wha J. Joo and Diana Shih and Richard C. Davis and Aldons J. Lusis and Eleazar Eskin, “ Meta-Analysis Identifies Gene-by-Environment Interactions as Demonstrated in a Study of 4,965 Mice ”, PLOS Genetics (2014) 10(1), e1004022. LINK

If you use our software to plot ForestPMPlot, please cite

Eun Yong Kang and Buhm Han and Eleazar Eskin, “ ForestPMplot: a flexible tool for visualizing heterogeneity between studies in meta analysis”, Submitted

If you refer multiple tissue eQTL method, please cite

Jae Hoon Sul and Buhm Han and Chun Ye and Ted Choi and Eleazar Eskin, “ Effectively identifying eQTLs from multiple tissues by combining mixed model and meta-analytic approaches.”, PLOS Genetics (2013) 9, e1003491. LINK Software

Contact

METASOFT

Buhm Han : buhan (AT) cs.ucla.edu , http:www.cs.ucla.du/~buhan

ForestPMPlot

Eun Yong Kang : ekang (AT) cs.ucla.edu

Funding information

B.H. and E.E. are supported by National Science Foundation grants 0513612, 0731455, 0729049 and 0916676, and NIH grants K25-HL080079 and U01-DA024417. B.H. is supported by the Samsung Scholarship. This research was supported in part by the University of California, Los Angeles subcontract of contract N01-ES-45530 from the National Toxicology Program and National Institute of Environmental Health Sciences to Perlegen Sciences.