Regional Heritability Analysis v1.0

Introduction

Regional Heritability Analysis is a software package for estimating regional heritability of small genomic segments. It uses LD score regression to estimate regional heritability of small genomic segments and examined if the heritability is evenly distributed across segments in each phenotype. Additionally, using the regional heritability data, we investigated relationship among the traits included in our analysis and searched for pleiotropic loci contributing much of heritability to many traits. For analysis result across all traits, visit our interactive database website (Regional Heritability Atlas, http://h2atlas.hanlab.snu.ac.kr).

Example

Nested donut plot

The estimated regional heritability over different scales (chromosome, 128Mb, 64Mb, 32Mb, 16Mb, and 8Mb) simultaneously shown in a nested donut plot. The chromosome to which each segment belongs are indicated by background color of the segment. Fraction of heritability which each segment explains can be measured using outer circular gauge.

Comparison of regional heritability between two traits

The chromosome to which each segment belongs are indicated by color of the segment.

Correlation between phenotypes calculated from regional heritability

Main functionality

This package provides full pipeline of data analysis from downloading data to visualizing results.

• Data pre-processing
  • download GWAS summary statistics from Neale lab
  • download raw genotype data of 1000 Genome phase3 in plink format
  • munge summary statistics (wrapper for munge_sumstats.py from https://github.com/bulik/ldsc)
  • calculate LD score (wrapper for ldsc.py --l2 from https://github.com/bulik/ldsc)
  • run LD score regression (wrapper for ldsc.py --h2 from https://github.com/bulik/ldsc)
• Analyze and visualize result
  • level of polygenicity
  • analysis of variance
  • nested donut plot
  • pleiotropic loci identification
  • correlation between phenotypes
• Build database website (demo: Regional Heritability Atlas, http://h2atlas.hanlab.snu.ac.kr)

Recommended (not necessary) directory structure

path can be configured in path_configure.py

├── data
│   ├── 1000G_Phase3_weights_hm3_no_MHC
│   ├── 1000G_plink_EUR
│   ├── 1000G_plink_EUR_temp
│   ├── hapmap3_snps
│   ├── out_sumstats
│   ├── out_annot
│   └── out_final
├── web
│   └── partitioned_heritability_website
│       └── plot_data
├── log_parser.py
├── path_configure.py
├── basic_tools.py
├── 1_downlaod_gwas_neale.ipynb
├── 2_munge_1000G_genotype.ipynb
├── 3_munge_sumstats.ipynb
├── 4_make_annot_ldscore.ipynb
├── 5_filtering_phenotypes_1.ipynb
├── 5_filtering_phenotypes_2.ipynb
├── 5_run_ldsc.ipynb
├── 6_saving_and_basic_qc.ipynb
├── 7_1_nested_donut_plot.ipynb
├── 7_2_anova.ipynb
├── 7_3_variance.ipynb
├── 7_4_correlation.ipynb
├── 7_5_pca.ipynb
├── 7_6_pleiotropic.ipynb
├── 7_7_alluvial.ipynb
├── 7_8_basic_qc.ipynb
├── 7_9_phenotype_info.ipynb
└── 8_bokeh.ipynb

Instructions

Environment

This software was tested on Linux (especially on CentOS 7).

Downloading the package

In order to download Regional Heritability Analysis, you can clone this repository.

$ git clone https://github.com/ch6845/regional_heritability_analysis.git
$ cd regional_heritability_analysis

Installing required dependencies

Some software packages must be installed.

• PLINK (download from https://www.cog-genomics.org/plink/1.9/)
  • make sure PLINK is added to the sysmtem path. verify it by plink --version
• Python 3.7.3 (downlod from https://www.python.org/ or https://www.anaconda.com/ and install following packages using pip install or conda install command)
  • Numpy 1.16.3
  • Pandas 0.24.2
  • Scikit-learn 0.21.2
  • SciPy 1.3.0
  • Matplotlib 3.0.3
  • Seaborn 0.9.0
  • Network 2.3,
• LDSC 1.0.0

For building web database,

• R 3.5.2
• Hugo 0.58.0
• Bokeh 1.3.4
• DataTables 1.10.19

Pipeline

1. 1_downlaod_gwas_neale.ipynb Download UK Biobank GWAS summary statistics from Neale lab
2. 2_munge_1000G_genotype.ipynb Prepare raw genotype of 1000 Genome project Phase 3 for building LD score reference panel
3. 3_munge_sumstats.ipynb Munge summary statistics
4. 4_make_annot_ldscore.ipynb Build LD score reference panel. Make use of its supporting bash shell argument . For example, run jupyter nbconvert 4_make_annot_ldscore.ipynb --to script and python 4_make_annot_ldscore.py 64 bp
5. Run LD score regression
   1. 5_filtering_phenotypes_1.ipynb Filter phenotypes satisfiying QC conditions
   2. 5_filtering_phenotypes_2.ipynb
   3. 5_run_ldsc.ipynb Build LD score reference panel. Make use of its supporting bash shell argument. For example, run jupyter nbconvert 5_run_ldsc.ipynb --to script and python 5_run_ldsc.py bp 64 10 0 500
6. 6_saving_and_basic_qc.ipynb Save result and check quaility
7. Analyze and visualizing result
   1. 7_1_nested_donut_plot.ipynb Nested donut plot
   2. 7_2_anova.ipynb Analysis of variance
   3. 7_3_variance.ipynb Level of polygenicity
   4. 7_4_correlation.ipynb Correlation between phenotypes included in analysis
   5. 7_5_pca.ipynb Principal component analysis
   6. 7_6_pleiotropic.ipynb Pleiotropic loci identification
   7. 7_7_alluvial.ipynb Region-wise comparision of regional heritiability of segments
   8. 7_8_basic_qc.ipynb Print quality check result
   9. 7_9_phenotype_info.ipynb Exporting analysis result
8. 8_bokeh.ipynb Build web database website automatically from analysis result

License

This project is licensed under the terms of the MIT license.

Citation

If you use the software Regional Heritability Analysis, please cite Kim and Han. Landscape of polygenicity of complex traits in UK Biobank. (under review) (2019)

References

1. PLINK v1.9 | Chang, Christopher C., et al. "Second-generation PLINK: rising to the challenge of larger and richer datasets." Gigascience 4.1 (2015): 7.
2. 1000 Genome Phase 3 data | 1000 Genomes Project Consortium. "A global reference for human genetic variation." Nature 526.7571 (2015): 68.
3. LD Score regression | Bulik-Sullivan, Brendan K., et al. "LD Score regression distinguishes confounding from polygenicity in genome-wide association studies." Nature genetics 47.3 (2015): 291.
4. Genetic map of GRCh build 37 | Howie, Bryan N., Peter Donnelly, and Jonathan Marchini. "A flexible and accurate genotype imputation method for the next generation of genome-wide association studies." PLoS genetics 5.6 (2009): e1000529.

To check full list of references, please refer to our publication.

Support

If you have any question, please feel free to contact us [email protected]