Compute JASS power gain from the genetic architecture of traits
In a recent study [SMenagerB+23], we explore how the genetic architecture of the set of traits (heritability, genetic covariance, heritability undetected by the univariate test, ...) can be predictive of statistical power gain of the multi-trait test.
We implement an additional command line tool to give access our predictive model (the jass predict-gain command). This command allows the to score swiftly a large number of traits combinations and to focus on set of traits the most promising for multi-trait testing.
To work the inittable provided to the jass predict-gain command must contain the genetic covariance between traits.
jass predict-gain --inittable-path inittable_curated_111_traits_20-03-2024.hdf5 --combination_path ./combination_example.tsv --gain-path predicted_gain.tsv
The second argument (--combination_path) is a path to a file containing the set of traits to be scored.
GRP1 |
z_GIANT_HIP z_GLG_HDL z_GLG_LDL z_MAGIC_2HGLU-ADJBMI |
|---|---|
GRP2 |
z_SPIRO-UKB_FVC z_SPIRO-UKB_FEV1 z_TAGC_ASTHMA |
When executed the command will created a report at --gain-path
traits |
k |
avg_distance_cor |
mean_gencov |
avg_Neff |
avg_h2 |
avg_perc_h2_diff_region |
log10_mean_gencov |
log10_avg_distance_cor |
gain |
|---|---|---|---|---|---|---|---|---|---|
['z_SPIRO-UKB_FVC'; 'z_SPIRO-UKB_FEV1'; 'z_TAGC_ASTHMA'] |
0.1 |
0.1731946683845993 |
0.0637 |
0.3843393026739591 |
0.2785193310634847 |
0.7976315890930669 |
0.8139196701681637 |
0.8013809378674498 |
0.06428524764535551 |
['z_GIANT_HIP'; 'z_GLG_HDL'; 'z_GLG_LDL'; 'z_MAGIC_2HGLU-ADJBMI'] |
0.2 |
0.14899001074867035 |
0.01535 |
0.12076877719858631 |
0.22628198390356655 |
0.9055326131023057 |
0.6573854616675169 |
0.7879956172999502 |
-0.010766494024690904 |
The last column provide the predicted gain ("the higher the more promising"). Note that extrapoling on new data might give lesser performances than reported in [SMenagerB+23].
Tomaz Berisa and Joseph K. Pickrell. Approximately independent linkage disequilibrium blocks in human populations. Bioinformatics, 32(2):283–285, 2015. doi:10.1093/bioinformatics/btv546.
Hanna Julienne, Vincent Laville, Zachary R McCaw, Zihuai He, Vincent Guillemot, Carla Lasry, Andrey Ziyatdinov, Cyril Nerin, Amaury Vaysse, Pierre Lechat, and others. Multitrait gwas to connect disease variants and biological mechanisms. PLoS genetics, 17(8):e1009713, 2021.
Hanna Julienne, Pierre Lechat, Vincent Guillemot, Carla Lasry, Chunzi Yao, Robinson Araud, Vincent Laville, Bjarni Vilhjalmsson, Hervé Ménager, and Hugues Aschard. Jass: command line and web interface for the joint analysis of gwas results. NAR Genomics and Bioinformatics, 2(1):lqaa003, 2020.
Hanna Julienne, Huwenbo Shi, Bogdan Pasaniuc, and Hugues Aschard. RAISS: robust and accurate imputation from summary statistics. Bioinformatics, 35(22):4837–4839, 06 2019. URL: https://doi.org/10.1093/bioinformatics/btz466, arXiv:https://academic.oup.com/bioinformatics/article-pdf/35/22/4837/30706731/btz466.pdf, doi:10.1093/bioinformatics/btz466.
Florian Privé. Optimal linkage disequilibrium splitting. Bioinformatics, 38(1):255–256, 07 2021. URL: https://doi.org/10.1093/bioinformatics/btab519, arXiv:https://academic.oup.com/bioinformatics/article-pdf/38/1/255/41891000/btab519.pdf, doi:10.1093/bioinformatics/btab519.