Sunbelt Computer Software

Heritability tools

Multiple tools calculate heritability, as shown in the table below. Each tool uses different statistical methods to estimate heritability, and the datasets they use also vary. Some tools use the GWAS summary statistic file, some use genotype data, covariates, and PCA, and some use reference panels and SNP tagging.

Kindly read this article for the statistical method and modeling assumptions used by each tool: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7330487/.

Tools

Tool	URL
GEMMA	https://github.com/genetics-statistics/GEMMA
GCTA	http://cnsgenomics.com/software/gcta/#Overview
LDAK	http://dougspeed.com/ldak/
DPR	https://github.com/biostatpzeng/DPR
LDSC	https://github.com/bulik/ldsc
SumHer	http://dougspeed.com/sumher/

Comparison of Methods for Heritability

This table presents a comparison of different methods used to calculate heritability in this research work. The first column shows the index of each method, the second column shows the tools used to calculate the heritability, and the third column shows the data and algorithm used to create a unique variant of each method. The GWAS, genotype data, covariates, and reference panel columns indicate whether specific data was used for the analysis or not. The last column indicates whether clumping or pruning was performed on the genotype data before it was used in the analysis. The final column shows the number of unique heritability values generated by each method.

Index	Tool	Models	GWAS	Genotype data	Covariates	Reference Panel	Clumping and Pruning	Models
Method 1	LDpred-2	LDpred-2_full	Yes	Yes	No	No	Yes/No	4
		LDpred-2_hapmap	Yes	Yes	No	No	Yes/No
Method 2	GCTA	(Data) Genotype	No	Yes	Yes	No	Yes/No	6
		Genotype+Covariate	No	Yes	Yes	No	Yes/No
		Genotype+Covariate+PCA	No	Yes	Yes	No	Yes/No
Method 3	GEMMA	(Related Matrix) Centered/ Standardized	No	Yes	Yes	No	Yes/No	24
		(Heritability Algorithm) HE regression	No	Yes	Yes	No	Yes/No
		REML AI algorithm	No	Yes	Yes	No	Yes/No
		(Data) Genotype	No	Yes	Yes	No	Yes/No
		Genotype+Covariate	No	Yes	Yes	No	Yes/No
		Genotype+Covariate+PCA	No	Yes	Yes	No	Yes/No
Method 4	GEMMA	(Heritability Algorithm) HE regression	Yes	Yes	Yes	No	Yes/No	6
		(Data) Genotype	Yes	Yes	Yes	No	Yes/No
		Genotype+Covariate	Yes	Yes	Yes	No	Yes/No
		Genotype+Covariate+PCA	Yes	Yes	Yes	No	Yes/No
Method 5	LDSC+GEMMA	(Heritability Algorithm) REML AI algorithm	Yes	Yes	Yes	LD Score of Genotype data	Yes	3
		(Data) Genotype	Yes	Yes	Yes	LD Score of Genotype data	Yes
		Genotype+Covariate	Yes	Yes	Yes	LD Score of Genotype data	Yes
		Genotype+Covariate+PCA	Yes	Yes	Yes	LD Score of Genotype data	Yes
Method 6	DPR+GEMMA	(Related Matrix) Centered/ Standardized - from DPR	No	Yes	Yes	No	Yes/No	24
		(Heritability Algorithm) HE regression	No	Yes	Yes	No	Yes/No
		REML AI algorithm	No	Yes	Yes	No	Yes/No
		(Data) Genotype	No	Yes	Yes	No	Yes/No
		Genotype+Covariate	No	Yes	Yes	No	Yes/No
		Genotype+Covariate+PCA	No	Yes	Yes	No	Yes/No
Method 7	LDSC	LDSC (EUR reference panel, EUR weights)	Yes	No	No	EUR reference panel EUR weights	No	4
Method 8	LDAK-Calculate	LDAK Tagging Models (Human, GCTA, BLD-LDAK, Alpha)	Yes	Yes	No	No	Yes/No	8
Method 9	LDAK-Precomputed	Precomputed taggings (bld.ldak.hapmap.gbr.tagging ldak.thin.hapmap.gbr.tagging bld.ldak.lite.alpha.hapmap.gbr.tagging)	Yes	No	No	No	No	6
		bld.ldak.genotyped.gbr.tagging	Yes	No	No	No	No
		ldak.thin.genotyped.gbr.tagging	Yes	No	No	No	No
		bld.ldak.lite.alpha.genotyped.gbr.tagging	Yes	No	No	No	No
Method 10	LDSC	LDSC (Reference panel, EUR weights from Genotype data)	Yes	Yes	No	LD Score of Genotype data	Yes	1

Result of heritability calculation for all 10 Phenotypes and 86 unique Methods.

Purpose of this documentation

In this research, we used various heritability tools and created multiple variants of each method to calculate heritability for 10 phenotypes. Two polygenic risk scores (PRS) tools, LDpred-2 and GCTA, rely on heritability estimates for PRS calculation. We investigated whether the method used to calculate heritability impacts the performance of the PRS tools. Benchmarking all these tools is essential to identify the best method for heritability calculation that optimizes PRS calculation.

Helper tools

Tool	Description	Link
GWASPokerforPRS	A tool for downloading GWAS data from GWAS Catalog	https://github.com/MuhammadMuneeb007/GWASPokerforPRS
Detect genomic build	Detect the genomic build of a dataset	https://www.biostars.org/p/9495682/#9595219
pyliftover	A Python package for genomic coordinate conversion	https://pypi.org/project/pyliftover/

Dataset

We analyzed 14 phenotypes from the UK Biobank and downloaded the corresponding GWAS files from the GWAS catalog (https://www.ebi.ac.uk/gwas/). After converting the genotype data to match the GWAS file's genotype build, we calculated the number of common variants between the GWAS files and the genotype data. Three phenotypes were removed from further analysis due to a limited number of variants.

GWAS Data Processing

Steps for Processing GWAS Data

Download the GWAS file for a specific phenotype.
Transform the GWAS file to a specific format accepted by most PRS tools.
- The sample transformation code for one phenotype (asthma) is shown below.

Original GWAS file for asthma:

chromosome	base_pair_location	effect_allele	other_allele	effect_allele_frequency	beta	standard_error	p_value	variant_id
1	100000012	T	G	0.28029314035229	0.0055	0.0077	0.4756	NA
1	10000006	A	G	0.00547967251882486	-0.1081	0.0560	0.05366	NA
1	100000135	T	A	0.00172932855852879	0.0255	0.1026	0.8037	NA
1	100000374	C	G	0.000580831145737331	0.0204	0.2956	0.9451	NA
1	100000827	T	C	0.329735132840309	0.0016	0.0074	0.8236	NA

Transformation Code:

import pandas as pd

df = pd.read_csv('gwas.csv.modified')
print(df.head())
print(df.columns)

column_map = {
    'chromosome': 'CHR', 
    'variant_id': 'SNP', 
    'base_pair_location': 'BP',
    'effect_allele': 'A1', 
    'other_allele': 'A2', 
    'p_value': 'P',
    'effect_allele_frequency': 'MAF', 
    'beta': 'BETA', 
    'standard_error': 'SE'
}

df = df.rename(columns=column_map)

df["SNP"] = "X"
df["INFO"] = 1
df["N"] = 449500

df = df[['CHR', 'BP', 'SNP', 'A1', 'A2', 'N', 'SE', 'P', 'BETA', 'INFO', 'MAF']]

df.to_csv('asthma.gz', compression='gzip', index=False, sep="\t")

CHR	BP	SNP	A1	A2	N	SE	P	BETA	INFO	MAF
1	100000012	X	T	G	449500	0.0077	0.4756	0.0055	1	0.28029314035229
1	10000006	X	A	G	449500	0.056	0.05366	-0.1081	1	0.0054796725188248
1	100000135	X	T	A	449500	0.1026	0.8037	0.0255	1	0.0017293285585287

3. Detect the genomic build of the GWAS.

4. Handle missing RSID or SNP information:

Generate two copies of the genotype data based on the GWAS build (hg19 or hg38).
Find the common predictors and use the SNPs in the genotype .bim file to fill the SNP column in the GWAS.

CHR	BP	SNP	A1	A2	N	SE	P	BETA	INFO	MAF
1	100008709	rs55727773	T	G	449500	0.0148	0.4387	-0.0115	1	0.881099991887257
1	100011667	rs113462541	C	T	449500	0.0096	0.4591	0.0071	1	0.835926208119123
1	100015703	rs12028261	T	C	449500	0.0071	0.6983	0.0027	1	0.505985657483862

5. Save the final file based on the following format:

# Define file paths for different data files
BED = asthma/asthma.bed
BIM = asthma/asthma.bim
FAM = asthma/asthma.fam
COV = asthma/asthma.COV
Height = asthma/asthma.PHENO
GWAS = asthma/asthma.gz

Below is the diagram showcasing data processing:

Processed Data

Phenotype	PMID	File Name	SNPs in GWAS (G)	SNPs in Genotype data (GE)	Common in G and GE	DOI	Cite
asthma	34594039	GCST90018795_buildGRCh37.tsv.gz	25837674	619653	2867	10.1038/s41588-021-00931-x	Sakaue2021
blood_pressure_medication	34662886	GCST90081464_buildGRCh38.tsv.gz	447993	619653	56	10.1038/s41586-021-04103-z	Backman2021
body_mass_index_bmi	34594039	GCST90018947_buildGRCh37.tsv.gz	20538803	619653	2866	10.1038/s41588-021-00931-x	Sakaue2021
cholesterol_lowering_medication	34662886	GCST90079486_buildGRCh38.tsv.gz	133986	619653	56	10.1038/s41586-021-04103-z	Backman2021
depression	29662059	UKBiobank_broad_12Jan18.txt	7641987	619653	545218	10.1038/s41467-018-03819-3	Howard2018
gastro_oesophageal_reflux_gord_gastric_reflux	34594039	GCST90018848_buildGRCh37.tsv.gz	25843128	619653	2906	10.1038/s41588-021-00931-x	Sakaue2021
hayfever_allergic_rhinitis	34662886	GCST90077815_buildGRCh38.tsv.gz	458874	619653	56	10.1038/s41586-021-04103-z	Backman2021
high_cholesterol	29892013	GCST90029021_buildGRCh37.tsv	12007882	619653	562302	10.1038/s41588-018-0144-6	Loh2018
hypertension	33893285	GCST90086092_buildGRCh37.tsv	15650645	619653	619653	10.1038/s41467-021-21952-4	GuindoMartinez2021
hypothyroidism_myxoedema	34594039	GCST90018862_buildGRCh37.tsv.gz	25801084	619653	2908	10.1038/s41588-021-00931-x	Sakaue2021
irritable_bowel_syndrome	34741163	GCST90016564_buildGRCh37.tsv	9885499	619653	568535	10.1038/s41588-021-00950-8	Eijsbouts2021
migraine	34737426	GCST90043745_buildGRCh37.tsv.gz	11831933	619653	548955	10.1038/s41588-021-00954-4	Jiang2021
osteoarthritis	36411363	GCST90134279_buildGRCh37.csv	24880768	619653	619653	10.1038/s41588-022-01221-w	McDonald2022

Perform Quality Controls on the Data

The initial quality controls vary from dataset to dataset and depend on how the dataset is being stored, processed, and handled.

Initial Directory Structure

Before executing Step1-MakeDirectories.py, you should have the following directory structure:

.
├── body_mass_index_bmi
│   ├── GCST90018947_buildGRCh37.tsv.gz
│   ├── gwas.csv.modified
│   └── Output.py
├── genotypes.bed
├── genotypes.bim19
├── genotypes.bim38
├── genotypes.fam
├── phenotype_file.txt
└── Step1-MakeDirectories.py

Execute the command below. It will check the build, and generate the files for a specific phenotype that includes covariate and bed bim and fam file. The code works for binary phenotype as well.

python Step1-MakeDirectories.py body_mass_index_bmi

Files after executing the above command.

.
├── body_mass_index_bmi
│   ├── body_mass_index_bmi.bed
│   ├── body_mass_index_bmi.bim
│   ├── body_mass_index_bmi.cov
│   ├── body_mass_index_bmi.covOLD
│   ├── body_mass_index_bmi.fam
│   ├── body_mass_index_bmi.gz
│   ├── body_mass_index_bmi.height
│   ├── body_mass_index_bmi.heightOLD
│   ├── GCST90018947_buildGRCh37.tsv.gz
│   ├── gwas.csv.modified
│   └── Output.py
├── genotypes.bed
├── genotypes.bim19
├── genotypes.bim38
├── genotypes.fam
├── phenotype_file.txt
└── Step1-MakeDirectories.py

Generating Multiple Folds and Performing Quality Controls

Match the SNPs between the GWAS and the genotype.bim file and update the RSID column.
For continuous phenotypes, employ normal cross-validation. For binary phenotypes, utilize stratified cross-validation with default train-test split, generating 5 folds.

Data Splitting and Organization

Split the data and save it in directories as outlined below:

Continuous Phenotype:
- Normal cross-validation (5 folds)
- Data saved in respective directories for each fold.
Binary Phenotype:
- Stratified cross-validation (5 folds)
- Data saved in separate directories for each fold.

Quality Controls

Perform quality controls on both the GWAS and the Genotype data, including:

Minor allele frequency adjustments
Duplicate SNP identification

SNP Matching

Use Module1.R to check common SNPs between the GWAS and the genotype data.
Rely on heritability tools for SNP matching.

python Step2-PerformQualityControls.py body_mass_index_bmi

Files after executing the above command.

.
├── body_mass_index_bmi
│   ├── body_mass_index_bmi.bed
│   ├── body_mass_index_bmi.bim
│   ├── body_mass_index_bmi.cov
│   ├── body_mass_index_bmi.covOLD
│   ├── body_mass_index_bmi.fam
│   ├── body_mass_index_bmi.gz
│   ├── body_mass_index_bmi.height
│   ├── body_mass_index_bmi.heightOLD
│   ├── body_mass_index_bmi_QC.bed
│   ├── body_mass_index_bmi_QC.bim
│   ├── body_mass_index_bmi_QC.cov
│   ├── body_mass_index_bmi_QC.fam
│   ├── body_mass_index_bmi_QC.log
│   ├── Fold_0
│   │   ├── test_data.bed
│   │   ├── test_data.bim
│   │   ├── test_data.cov
│   │   ├── test_data.fam
│   │   ├── test_data.log
│   │   ├── train_data.a1
│   │   ├── train_data.bed
│   │   ├── train_data.bim
│   │   ├── train_data.cov
│   │   ├── train_data.fam
│   │   ├── train_data.log
│   │   ├── train_data.mismatch
│   │   ├── train_data.QC.bed
│   │   ├── train_data.QC.bim
│   │   ├── train_data.QC.fam
│   │   ├── train_data.QC.het
│   │   ├── train_data.QC.log
│   │   ├── train_data.QC.rel.id
│   │   ├── train_data.QC.snplist
│   │   └── train_data.valid.sample
│   ├── Fold_1
│   ├── Fold_2
│   ├── Fold_3
│   ├── Fold_4
│   ├── GCST90018947_buildGRCh37.tsv.gz
│   ├── gwas.csv.modified
│   ├── Output.py
│   └── PeopleWithPhenotype.txt
├── genotypes.bed
├── genotypes.bim19
├── genotypes.bim38
├── genotypes.fam
├── Module1.R
├── phenotype_file.txt
├── plink
├── Step1-MakeDirectories.py
└── Step2-PerformQualityControls.py

For all methods, we considered the clumping and pruning parameters as shown below.

Parameter	Value
clump_p1	1
clump_r2	0.1
clump_kb	200
p_window_size	200
p_slide_size	50
p_LD_threshold	0.25
numberofpca	6

Method 1

Once the data is processed, we can calculate heritability using LDpred-2.

For a fold 0, execute the specific method:

python Method1.py body_mass_index_bmi 0

See Method1.md for the detailed results

Method 2

To calculate h2 using GCTA, we followed the GCTA tutorial.

Using only genotype data and phenotype. The name in the code is GCTA_genotype.
- gcta --bfile FILE --make-grm --out FILE
- gcta --grm FILE --pheno FILE.PHENP --reml --out FILE
Using genotype data, covariates, and phenotype. The name in the code is GCTA_genotype_covariate.
- gcta --bfile FILE --make-grm --out FILE
- gcta --grm FILE --pheno FILE.PHENP --qcovar FILE.cov --reml --out FILE
Using genotype data, covariates, PCA, and phenotype. The name in the code is GCTA_genotype_covariate_pca.
- gcta --bfile FILE --make-grm --out FILE
- gcta --grm FILE --pheno FILE.PHENP --reml --qcovar FILE.cov_pca --out FILE

OR

- `gcta --grm FILE --pheno FILE.PHENP --reml-no-constrain --qcovar  FILE.cov_pca --out FILE`

Handling REML Non-convergence:

In some cases, the REML calculation may not converge (source). To address this, the --reml-no-constrain flag can be used. However, it's important to note that in such cases, the heritability value might exceed 1, often attributed to sample relatedness (source%20Heritability%20can%20be%20greater,can%20also%20cause%20this%20result.)).

An alternative approach involves using a flag like --grm-cutoff 0.025, similar to Plink's --rel-cutoff 0.125. This allows for the exclusion of specific samples, facilitating the recalculation of heritability.

gcta --grm FILE --pheno FILE.PHENP --reml --grm-cutoff 0.025 --qcovar FILE.cov_pca --out FILE

Summary result of REML analysis:

h2 = V(G)/Vp

Once the data is processed, we can calculate heritability using GCTA.

For a fold 0, execute the specific method:

python Method2.py body_mass_index_bmi 0

See Method2.md for the detailed results

Method 3

GEMMA employs various methods for calculating heritability. For details on these calculations, please refer to their manual.

./gemma --bfile BFILE -gk relatedmatrix -o output
./gemma --bfile train_data -gk 1 -o output

 
./gemma -p BFILE.fam -k output/output.cXX.txt -n 6 -vc model -o output
./gemma -p train_data.fam -k output/output.cXX.txt -n 6 -vc 1 -o output

See Method3.md for the detailed results

Method 4

GEMMA employs various methods for calculating heritability. For details on these calculations, please refer to their manual.

./gemma -beta GWAS.txt -bfile GENOTYPE -vc models -c COV.COV_PCAgemma -o output

See Method4.md for the detailed results

Method 5

In this method, we first used the genotype data and performed LD score regression analysis, and that file was further used by Gemma for calculating heritability. The LDSC parameters can be changed.

Commands

python ldsc.py --bfile BFILE --yes-really --l2 --ld-wind-cm 1 --out traindirec/ld
./gemma -beta GWAS -bfile BFILE -wcat traindirec/ld -vc 2 -o OUT

See Method5.md for the detailed results

Method 6

In this method, we used DPR + GEMMA. First, we used DPR to calculate the related matrix that is centered and standardized from DPR. The heritability is estimated using the HE regression or REML AI algorithm. The data can include genotype only, genotype with covariates, or genotype with covariates and PCA.

./DPR --bfile BFILE -gk 1 or 2 -o OUT
./gemma -p BFILE.fam -k OUT -n 6 -vc model -c COV_PCAgemma -o OUT2

See Method6.md for the detailed results

Method 7

LDSC provides a simple way to calculate heritability using the GWAS file. Follow their tutorial for detailed calculation.

LDSC Heritability and Genetic Correlation

Important Files:

ref-ld-chr: Reference file (eur_ref_ld_chr.tar.bz2)
w-ld-chr: Regression weights (eur_w_ld_chr.tar.bz2)
w_hm3.snplist: SNP list file (w_hm3.snplist.bz2)

python ldsc.py --h2 GWAS --ref-ld-chr REF --w-ld-chr WEIGHT --out OUT

See Method7.md for the detailed results

Method 8

LDAK

LDAK calculates heritability using SNP tagging. Heritability using LDAK can be calculated using the precomputed taggings and by calculating SNPs tagging from the genotype data.

LDAK provides four different models to calculate taggings: "human", "GCTA", "BLD-LDAK", "alpha". These taggings are then combined with the GWAS file to calculate heritability. The tutorial they provided is pretty good and provides a detailed explanation for each step.

./ldak --calc-tagging TAGGING --bfile GENOTYPE --power -.25

./ldak --sum-hers OUTPUT_FILENAME --summary GWAS --tagfile TAGGING --check-sums NO

See Method8.md for the detailed results

Method 9

LDAK also provides precomputed taggings which were calculated from the UK Biobank data. They also provided taggings for different populations, which were generated using different data sources from different models.

Tagging Files

bld.ldak.hapmap.gbr.tagging
ldak.thin.hapmap.gbr.tagging
bld.ldak.lite.alpha.hapmap.gbr.tagging
bld.ldak.genotyped.gbr.tagging
ldak.thin.genotyped.gbr.tagging
bld.ldak.lite.alpha.genotyped.gbr.tagging

./ldak --sum-hers OUTPUT_FILENAME --summary GWAS --tagfile your_precomputed_file --check-sums NO

See Method9.md for the detailed results

Method 10

LDSC also provides a way to calculate the reference panel from the genotype data and then combine the genotype data and the GWAS file to calculate the heritability. Follow the LDSC LD Score Estimation Tutorial for detailed instructions.

Create Binary Fileset for Each Chromosome Using PLINK
```
./plink --bfile BFILE --chr chr_num --make-bed --out BFILE_chr_num
```
- Command: ./plink --bfile BFILE --chr chr_num --make-bed --out BFILE_chr_num
- Description: This command uses PLINK to create a binary fileset (.bed, .bim, .fam) for a specific chromosome (chr_num) from the base file (BFILE). The output is saved with a suffix indicating the chromosome number.
Calculate LD Scores Using LDSC
```
python ldsc.py --bfile BFILE_chr_num --l2 --yes-really --ld-wind-cm 1 --out LDSCFILES/chr_num
```
- Command: python ldsc.py --bfile BFILE_chr_num --l2 --yes-really --ld-wind-cm 1 --out LDSCFILES/chr_num
- Description: This command runs LDSC to calculate LD scores for the chromosome-specific binary fileset created by PLINK. It uses a window size of 1 centimorgan and outputs the results to the specified directory (LDSCFILES/chr_num).
Estimate Heritability Using LDSC
```
python ldsc.py --h2 GWAS --ref-ld-chr LDSCFILES/ --w-ld-chr LDSCFILES/ --out OUT
```
- Command: python ldsc.py --h2 GWAS --ref-ld-chr LDSCFILES/ --w-ld-chr LDSCFILES/ --out OUT
- Description: This command estimates heritability using the specified GWAS summary statistics file (GWAS). It uses reference LD scores and weights from the specified directories (LDSCFILES/) and outputs the results to the specified directory (OUT).

See Method10.md for the detailed results

Name		Name	Last commit message	Last commit date
Latest commit History 71 Commits
All.csv		All.csv
All_heritability.csv		All_heritability.csv
Data.png		Data.png
FoldParameters.txt		FoldParameters.txt
LDpred-2-Heritability.R		LDpred-2-Heritability.R
Method1.md		Method1.md
Method1.py		Method1.py
Method1.sh		Method1.sh
Method10.md		Method10.md
Method10.py		Method10.py
Method10.sh		Method10.sh
Method2.md		Method2.md
Method2.py		Method2.py
Method2.sh		Method2.sh
Method3.md		Method3.md
Method3.py		Method3.py
Method3.sh		Method3.sh
Method4.md		Method4.md
Method4.py		Method4.py
Method4.sh		Method4.sh
Method5.md		Method5.md
Method5.py		Method5.py
Method5.sh		Method5.sh
Method6.md		Method6.md
Method6.py		Method6.py
Method6.sh		Method6.sh
Method7.md		Method7.md
Method7.py		Method7.py
Method7.sh		Method7.sh
Method8-Calculated.py		Method8-Calculated.py
Method8-Precomputed.py		Method8-Precomputed.py
Method8.md		Method8.md
Method8.py		Method8.py
Method8.sh		Method8.sh
Method9.md		Method9.md
Method9.py		Method9.py
Method9.sh		Method9.sh
Module1.R		Module1.R
PhenotypeParameters.txt		PhenotypeParameters.txt
README.md		README.md
Result2.png		Result2.png
Result_distribution.png		Result_distribution.png
Step1-MakeDirectories.py		Step1-MakeDirectories.py
Step2-PerformQualityControls.py		Step2-PerformQualityControls.py
Suppmenentary Material 1.xlsx		Suppmenentary Material 1.xlsx
gene_enrich.yml		gene_enrich.yml
heritability_availability_heatmap.png		heritability_availability_heatmap.png
heritability_supplementary.csv		heritability_supplementary.csv
ldscc.yml		ldscc.yml

Source	Variance	SE
V(G)	0.869695	0.790524
V(e)	0.000001	0.787625
Vp	0.869696	0.063776
V(G)/Vp	0.999999	0.905632

Sunbelt Computer Software

PL/B Language Development and Support

Folders and files

Latest commit

History

Repository files navigation

Heritability tools

Tools

Comparison of Methods for Heritability

Result of heritability calculation for all 10 Phenotypes and 86 unique Methods.

Purpose of this documentation

Helper tools

Dataset

GWAS Data Processing

Steps for Processing GWAS Data

Original GWAS file for asthma:

Processed Data

Perform Quality Controls on the Data

Initial Directory Structure

Generating Multiple Folds and Performing Quality Controls

Data Splitting and Organization

Quality Controls

SNP Matching

Method 1

Method 2

Method 3

Method 4

Method 5

Commands

Method 6

Method 7

Important Files:

Method 8

LDAK

Method 9

Tagging Files

Method 10

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Uh oh!

Contributors

Uh oh!

Languages

Packages