SuShiE (Sum of Shared Single Effect) is a Python software to fine-map causal SNPs, compute prediction weights, and infer effect size correlation for molecular data (e.g., mRNA levels and protein levels etc.) across multiple ancestries.
- We detest usage of our software or scientific outcome to promote racial discrimination.SuShiE is described in
Zeyun Lu, Xinran Wang, Matthew Carr, Artem Kim, Steven Gazal, Pejman Mohammadi, Lang Wu, Alexander Gusev, James Pirruccello, Linda Kachuri, Nicholas Mancuso.
Check here for full documentation.
Installation | Example | Notes | Version History | Support | Other Software
Users can download the latest repository and then use pip:
git clone https://github.com/mancusolab/sushie.git
cd sushie
pip install .We currently only support Python3.8+.
Before installation, we recommend to create a new environment using conda so that it will not affect the software versions of the other projects.
SuShiE software is very easy to use:
For fine-mapping using individual-level data:
cd ./data/
sushie finemap --pheno EUR.pheno AFR.pheno --vcf vcf/EUR.vcf vcf/AFR.vcf --covar EUR.covar AFR.covar --output ./test_resultFor fine-mapping using summary-level data:
cd ./data/
sushie finemap --summary --gwas EUR.gwas AFR.gwas --vcf vcf/EUR.vcf vcf/AFR.vcf --sample-size 489 639 --gwas-header chrom snp pos a1 a0 zs --output ./test_resultIt can perform:
- SuShiE: multi-ancestry fine-mapping accounting for ancestral correlation
- Single-ancestry SuSiE (Sum of Single Effect)
- Independent SuShiE: multi-ancestry SuShiE without accounting for correlation
- Meta-SuSiE: single-ancestry SuSiE followed by meta-analysis
- Mega-SuSiE: single-ancestry SuSiE on row-wise stacked data across ancestries (individual-level data only)
- cis-molQTL effect size correlation estimation
- cis-SNP heritability estimation (individual-level data only)
- Cross-validation for SuShiE prediction weights (individual-level data only)
- Convert prediction results to FUSION format, thus can be used in TWAS (individual-level data only)
See here for more details on how to use SuShiE.
If you want to use in-software SuShiE inference function, you can use following Python code as an example:
from sushie.infer import infer_sushie
# Xs is for genotype data, and it should be a list of numpy array whose length is the number of ancestry.
# ys is for phenotype data, and it should also be a list of numpy array whose length is the number of ancestry.
infer_sushie(Xs=X, ys=y)
# Or summary-level data
# lds is for LD data, and it should be a list of p by p numpy array whose length is the number of ancestry.
# zs is for GWAS data, and it should be a list of numpy array whose length is the number of ancestry/
infer_sushie_ss(lds=LD, zs=GWAS, ns=np.array([100, 100]))You can customize this function with your own ideas!
- SuShiE currently only supports continuous phenotype fine-mapping for individual-level data.
- SuShiE uses JAX with Just In
Time
compilation to achieve high-speed computation. However, there are
some issues for JAX
with Mac M1 chip. To solve this, users need to initiate conda using
miniforge, and then
install SuShiE using
pipin the desired environment.
| Version | Description |
|---|---|
| 0.1 | Initial Release |
| 0.11 | Fix the bug for OLS to compute adjusted r squared. |
| 0.12 | Update io.corr function so that report all the correlation results no matter cs is pruned or not. |
| 0.13 | Add --keep command to enable user to specify a file that contains the subjects ID SuShiE will perform on. Add --ancestry_index command to enable user to specify a file that contains the ancestry index for fine-mapping. With this, user can input single phenotype, genotype, and covariate file that contains all the subjects across ancestries. Implement padding to increase inference time. Record elbo at each iteration and can access it in the infer.SuShiEResult object. The alphas table now outputs the average purity and KL divergence for each L. Change --kl_threshold to --divergence. Add --maf command to remove SNPs that less than minor allele frequency threshold within each ancestry. Add --max_select command to randomly select maximum number of SNPs to compute purity to avoid unnecessary memory spending. Add a QC function to remove duplicated SNPs. |
| 0.14 | Remove KL-Divergence pruning. Enhance command line appearance and improve the output files contents. Fix small bugs on multivariate KL. |
| 0.15 | Fix several typos; add a sanity check on reading vcf genotype data by assigning gt_types==Unknown as NA; Add preprint information. |
| 0.16 | Implement summary-level data inference. Add option to remove ambiguous SNPs; fix several bugs and enhance codes quality. |
For any questions, comments, bug reporting, and feature requests, please contact Zeyun Lu (zeyunlu@usc.edu) and Nicholas Mancuso (nmancuso@usc.edu), and open a new thread in the Issue Tracker.
Feel free to use other software developed by Mancuso Lab:
- MA-FOCUS: a Bayesian fine-mapping framework using TWAS statistics across multiple ancestries to identify the causal genes for complex traits.
- SuSiE-PCA: a scalable Bayesian variable selection technique for sparse principal component analysis
- twas_sim: a Python software to simulate TWAS statistics.
- FactorGo: a scalable variational factor analysis model that learns pleiotropic factors from GWAS summary statistics.
- HAMSTA: a Python software to estimate heritability explained by local ancestry data from admixture mapping summary statistics.
- Traceax: a Python library to perform stochastic trace estimation for linear operators.
This project has been set up using PyScaffold 4.1.1. For details and usage information on PyScaffold see https://pyscaffold.org/.