mvSuSiE RSS ukb

Last updated: 2021-04-05

Checks: 7 0

Knit directory: mmbr-rss-dsc/

This reproducible R Markdown analysis was created with workflowr (version 1.6.2). The Checks tab describes the reproducibility checks that were applied when the results were created. The Past versions tab lists the development history.

R Markdown file: up-to-date

Great! Since the R Markdown file has been committed to the Git repository, you know the exact version of the code that produced these results.

Environment: empty

Great job! The global environment was empty. Objects defined in the global environment can affect the analysis in your R Markdown file in unknown ways. For reproduciblity it’s best to always run the code in an empty environment.

Seed: set.seed(20200227)

The command set.seed(20200227) was run prior to running the code in the R Markdown file. Setting a seed ensures that any results that rely on randomness, e.g. subsampling or permutations, are reproducible.

Session information: recorded

Great job! Recording the operating system, R version, and package versions is critical for reproducibility.

Cache: none

Nice! There were no cached chunks for this analysis, so you can be confident that you successfully produced the results during this run.

File paths: relative

Great job! Using relative paths to the files within your workflowr project makes it easier to run your code on other machines.

Repository version: 5447dc7

Great! You are using Git for version control. Tracking code development and connecting the code version to the results is critical for reproducibility.

The results in this page were generated with repository version 5447dc7. See the Past versions tab to see a history of the changes made to the R Markdown and HTML files.

Note that you need to be careful to ensure that all relevant files for the analysis have been committed to Git prior to generating the results (you can use wflow_publish or wflow_git_commit). workflowr only checks the R Markdown file, but you know if there are other scripts or data files that it depends on. Below is the status of the Git repository when the results were generated:


Ignored files:
    Ignored:    .DS_Store
    Ignored:    .Rhistory
    Ignored:    .Rproj.user/
    Ignored:    data/.DS_Store
    Ignored:    output/.DS_Store

Untracked files:
    Untracked:  data/ENSG00000140265.12.Multi_Tissues.rds
    Untracked:  data/FastQTLSumStats.mash.FL_PC3.rds
    Untracked:  data/Multi_Tissues.ENSG00000089486.RDS
    Untracked:  data/mmbr_ukb_simulation/
    Untracked:  data/mmbr_ukb_simulation_problem/
    Untracked:  data/ukb_rss_naive_lfsr_problem.rds
    Untracked:  data/ukb_rss_naive_pip_problem.rds
    Untracked:  output/GTExprofile_res.rds
    Untracked:  output/GTExprofile_resL1.rds
    Untracked:  output/GTExprofile_resL1_elbo.rds
    Untracked:  output/GTExprofile_resL3.rds
    Untracked:  output/GTExprofile_resL3_elbo.rds
    Untracked:  output/GTExprofile_res_elbo.rds
    Untracked:  output/GTExprofile_resapprox.rds
    Untracked:  output/GTExprofile_resapproxL1.rds
    Untracked:  output/GTExprofile_resapproxL1_elbo.rds
    Untracked:  output/GTExprofile_resapproxL3.rds
    Untracked:  output/GTExprofile_resapproxL3_elbo.rds
    Untracked:  output/GTExprofile_resapprox_elbo.rds
    Untracked:  output/GTExprofile_resapproxdiag.rds
    Untracked:  output/GTExprofile_resapproxdiagL1.rds
    Untracked:  output/GTExprofile_resapproxdiagL1_elbo.rds
    Untracked:  output/GTExprofile_resapproxdiagL3.rds
    Untracked:  output/GTExprofile_resapproxdiagL3_elbo.rds
    Untracked:  output/GTExprofile_resapproxdiag_elbo.rds
    Untracked:  output/GTExprofile_resdiag.rds
    Untracked:  output/mmbr_gtex_res.Rprof
    Untracked:  output/mmbr_gtex_res_approx.Rprof
    Untracked:  output/mmbr_gtex_res_approx_diag.Rprof
    Untracked:  output/mmbr_gtex_res_diag.Rprof
    Untracked:  output/mnm_missing_output.20200527.rds
    Untracked:  output/test
    Untracked:  output/tiny_data_211_cond2L2.gif
    Untracked:  output/tiny_data_211_cond2L2.pdf
    Untracked:  output/tiny_data_211_cond2L3.gif
    Untracked:  output/tiny_data_211_cond2L3.pdf
    Untracked:  output/tiny_data_211_cond2initL3.gif
    Untracked:  output/tiny_data_211_cond2initL3.pdf
    Untracked:  output/ukb_rss_20200107/
    Untracked:  output/ukb_rss_20210313/

Unstaged changes:
    Modified:   analysis/mmbr_missing_rss_problem1.Rmd

Note that any generated files, e.g. HTML, png, CSS, etc., are not included in this status report because it is ok for generated content to have uncommitted changes.

These are the previous versions of the repository in which changes were made to the R Markdown (analysis/mmbr_rss_ukb_simulation_20210313.Rmd) and HTML (docs/mmbr_rss_ukb_simulation_20210313.html) files. If you’ve configured a remote Git repository (see ?wflow_git_remote), click on the hyperlinks in the table below to view the files as they were in that past version.

File	Version	Author	Date	Message
Rmd	5447dc7	zouyuxin	2021-04-05	wflow_publish(“analysis/mmbr_rss_ukb_simulation_20210313.Rmd”)
html	947ec35	zouyuxin	2021-03-16	Build site.
Rmd	99eed13	zouyuxin	2021-03-16	wflow_publish(“analysis/mmbr_rss_ukb_simulation_20210313.Rmd”)
html	38a8e4e	zouyuxin	2021-03-16	Build site.
Rmd	a7a9c47	zouyuxin	2021-03-16	wflow_publish(“analysis/mmbr_rss_ukb_simulation_20210313.Rmd”)

This is result from our mvSuSiE RSS simulation using UKB data. There are 600 datasets. The max PVE across traits is 0.0005.

For each dataset, we simulate signals using 2 type of priors, the details are here

Artificial mixture: 20 conditions. The oracle residual variance is a diagonal matrix.
UKB Bloodcells mixture: 16 conditions. The oracle residual variance is a dense matrix.

We estimate prior weights using ‘EM’ method.

Comparing with previous simulation 20210107, we add a small diagonal to ED priors.

Global comparison

We compare the CS and PIP for each SNP.

PIP calibration

Artificial Mixture

Different residuals with Oracle prior

Different priors with Oracle residual

Different priors with estimated residual from close to null z scores

UKB Bloodcells Mixture

Different residuals with Oracle prior

Different priors with Oracle residual

Different priors with estimated residual from close to null z scores

Power

Artificial Mixture

Different residuals with Oracle prior

Different priors with Oracle residual

Different priors with estimated residual from Y

Different priors with estimated residual from close to null z scores

UKB Bloodcells Mixture

Different residuals with Oracle prior

Different priors with Oracle residual

Different priors with estimated residual from Y

Different priors with estimated residual from close to null z scores

CS

Artificial Mixture

library(dplyr)


Attaching package: 'dplyr'

The following objects are masked from 'package:stats':

    filter, lag

The following objects are masked from 'package:base':

    intersect, setdiff, setequal, union

library(kableExtra)


Attaching package: 'kableExtra'

The following object is masked from 'package:dplyr':

    group_rows

tb = readRDS('output/ukb_rss_20210313/ukb_rss_20210313_cs/ukb_rss_cs_simuartificial_mixture_ukb_glob.rds')
tb$method = rownames(tb)
rename = list('susie_suff+FALSE' = 'SuSiE',
              'susie_rss+FALSE' = 'SuSiE-RSS',
              'mnm_suff_oracle+oracle' = 'mvSuSiE Oracle prior Oracle residual',
              'mnm_suff_oracle+covY' = 'mvSuSiE Oracle prior Y residual',
              'mnm_suff_identity+oracle' = 'mvSuSiE Random effects prior Oracle residual',
              'mnm_suff_identity+covY' = 'mvSuSiE Random effects prior Y residual',
              'mnm_suff_naive+oracle' = 'mvSuSiE Default prior Oracle residual',
              'mnm_suff_naive+covY' = 'mvSuSiE Default prior Y residual',
              'mnm_suff_ed+oracle' = 'mvSuSiE ED prior Oracle residual',
              'mnm_suff_ed+covY' = 'mvSuSiE ED prior Y residual',
              'mnm_suff_ed_ddcan+oracle' = 'mvSuSiE ED+default prior Oracle residual',
              'mnm_suff_ed_ddcan+covY' = 'mvSuSiE ED+default prior Y residual',
              'mnm_rss_oracle+oracle' = 'mvSuSiE-RSS Oracle prior Oracle residual',
              'mnm_rss_oracle+identity' = 'mvSuSiE-RSS Oracle prior Identity residual',
              'mnm_rss_oracle+nullz' = 'mvSuSiE-RSS Oracle prior z residual', 
              'mnm_rss_oracle+corY' = 'mvSuSiE-RSS Oracle prior Y residual',
              'mnm_rss_identity+oracle' = 'mvSuSiE-RSS Random effects prior Oracle residual',
              'mnm_rss_shared+oracle' = 'mvSuSiE-RSS Fixed effects prior Oracle residual',
              'mnm_rss_naive+oracle' = 'mvSuSiE-RSS Default prior Oracle residual',
              'mnm_rss_ed+oracle' = 'mvSuSiE-RSS ED prior Oracle residual',
              'mnm_rss_ed_ddcan+oracle' = 'mvSuSiE-RSS ED+default prior Oracle residual',
              'mnm_rss_identity_corY+corY' = 'mvSuSiE-RSS Random effects prior Y residual',
              'mnm_rss_shared_corY+corY' = 'mvSuSiE-RSS Fixed effects prior Y residual',
              'mnm_rss_naive_corY+corY' = 'mvSuSiE-RSS Default prior Y residual',
              'mnm_rss_ed_corY+corY' = 'mvSuSiE-RSS ED prior Y residual',
              'mnm_rss_ed_ddcan_corY+corY' = 'mvSuSiE-RSS ED+default prior Y residual',
              'mnm_rss_identity_corZ+nullz' = 'mvSuSiE-RSS Random effects prior z residual',
              'mnm_rss_shared_corZ+nullz' = 'mvSuSiE-RSS Fixed effects prior z residual',
              'mnm_rss_naive_corZ+nullz' = 'mvSuSiE-RSS Default prior z residual',
              'mnm_rss_ed_corZ+nullz' = 'mvSuSiE-RSS ED prior z residual',
              'mnm_rss_ed_ddcan_corZ+nullz' = 'mvSuSiE-RSS ED+default prior z residual')

Different residuals with Oracle prior

methods_resid = c('mnm_suff_oracle+oracle','mnm_rss_oracle+oracle', 
                  'mnm_rss_oracle+identity','mnm_rss_oracle+nullz','mnm_rss_oracle+corY')
rates_resid = tb %>% filter(method %in% methods_resid)
rates_resid$method = sapply(rates_resid$method, function(x) rename[[x]])
rates_resid$method = gsub(' Oracle prior', '', rates_resid$method)
rates_resid = rates_resid[match(c('mvSuSiE Oracle residual','mvSuSiE-RSS Oracle residual','mvSuSiE-RSS Identity residual',
                                  'mvSuSiE-RSS Y residual','mvSuSiE-RSS z residual'), rates_resid$method),]
rates_resid[,c('size', 'purity', 'overlap', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	power	fdr	coverage
mnm_suff_oracle+oracle	2	0.9993099	0.7217940	0.0246212	0.9753788
mnm_rss_oracle+oracle	2	0.9993099	0.7217940	0.0246212	0.9753788
mnm_rss_oracle+identity	2	0.9993172	0.7224947	0.0245979	0.9754021
mnm_rss_oracle+corY	2	0.9993008	0.7203924	0.0246679	0.9753321
mnm_rss_oracle+nullz	2	0.9992816	0.7189909	0.0265655	0.9734345

Different priors with Oracle residual

methods_prior_oracleresid = c('mnm_rss_oracle+oracle', 'mnm_rss_identity+oracle', 'mnm_rss_shared+oracle',
                              'mnm_rss_naive+oracle', 'mnm_rss_ed+oracle','mnm_rss_ed_ddcan+oracle')
rates_priors_oracleresid = tb %>% filter(method %in% methods_prior_oracleresid)
rates_priors_oracleresid$method = sapply(rates_priors_oracleresid$method, function(x) rename[[x]])
rates_priors_oracleresid$method = gsub(' Oracle residual', '', rates_priors_oracleresid$method)
rates_priors_oracleresid = rates_priors_oracleresid[match(c('mvSuSiE-RSS Oracle prior','mvSuSiE-RSS Fixed effects prior',
                                                            'mvSuSiE-RSS Random effects prior','mvSuSiE-RSS Default prior',
                                                            'mvSuSiE-RSS ED prior','mvSuSiE-RSS ED+default prior'), 
                                                          rates_priors_oracleresid$method),]
rates_priors_oracleresid[,c('size', 'purity', 'overlap', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	overlap	power	fdr	coverage
mnm_rss_oracle+oracle	2	0.9993099	0	0.7217940	0.0246212	0.9753788
mnm_rss_shared+oracle	2	0.9976247	0	0.3615978	0.0318949	0.9681051
mnm_rss_identity+oracle	2	0.9990295	0	0.6944639	0.0284314	0.9715686
mnm_rss_naive+oracle	2	0.9989653	35	0.7231955	0.0227273	0.9772727
mnm_rss_ed+oracle	2	0.9991060	0	0.7168886	0.0266413	0.9733587
mnm_rss_ed_ddcan+oracle	2	0.9992153	4	0.7189909	0.0247148	0.9752852

Different priors with estimated residual from Y

methods_prior_y = c('mnm_rss_oracle+corY', 'mnm_rss_identity_corY+corY',
                        'mnm_rss_shared_corY+corY', 'mnm_rss_naive_corY+corY',
                        "mnm_rss_ed_corY+corY","mnm_rss_ed_ddcan_corY+corY")
rates_priors_yresid = tb %>% filter(method %in% methods_prior_y)
rates_priors_yresid$method = sapply(rates_priors_yresid$method, function(x) rename[[x]])
rates_priors_yresid$method = gsub(' Y residual', '', rates_priors_yresid$method)
rates_priors_yresid = rates_priors_yresid[match(c('mvSuSiE-RSS Oracle prior','mvSuSiE-RSS Fixed effects prior',
                                                  'mvSuSiE-RSS Random effects prior','mvSuSiE-RSS Default prior',
                                                  'mvSuSiE-RSS ED prior','mvSuSiE-RSS ED+default prior'), 
                                                rates_priors_yresid$method),]
rates_priors_yresid[,c('size', 'purity', 'overlap', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	overlap	power	fdr	coverage
mnm_rss_oracle+corY	2	0.9993008	0	0.7203924	0.0246679	0.9753321
mnm_rss_shared_corY+corY	2	0.9975845	0	0.3622985	0.0318352	0.9681648
mnm_rss_identity_corY+corY	2	0.9990577	0	0.6937631	0.0294118	0.9705882
mnm_rss_naive_corY+corY	2	0.9989653	35	0.7210932	0.0237192	0.9762808
mnm_rss_ed_corY+corY	2	0.9991516	0	0.7168886	0.0266413	0.9733587
mnm_rss_ed_ddcan_corY+corY	2	0.9992418	4	0.7189909	0.0247148	0.9752852

Different priors with estimated residual from close to null z scores

methods_prior = c('mnm_rss_oracle+nullz', 'mnm_rss_identity_corZ+nullz', 
                  'mnm_rss_shared_corZ+nullz', 'mnm_rss_naive_corZ+nullz',
                  'mnm_rss_ed_corZ+nullz', 'mnm_rss_ed_ddcan_corZ+nullz')
rates_priors = tb %>% filter(method %in% methods_prior)
rates_priors$method = sapply(rates_priors$method, function(x) rename[[x]])
rates_priors$method = gsub(' Y residual', '', rates_priors$method)
rates_priors$method = gsub(' z residual', '', rates_priors$method)
rates_priors = rates_priors[match(c('mvSuSiE-RSS Oracle prior', 'mvSuSiE-RSS Random effects prior',
                                    'mvSuSiE-RSS Fixed effects prior','mvSuSiE-RSS Default prior',
                                    'mvSuSiE-RSS ED prior','mvSuSiE-RSS ED+default prior'), rates_priors$method),]
rates_priors[,c('size', 'purity', 'overlap', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	overlap	power	fdr	coverage
mnm_rss_oracle+nullz	2	0.9992816	0	0.7189909	0.0265655	0.9734345
mnm_rss_identity_corZ+nullz	2	0.9991516	0	0.6951647	0.0340798	0.9659202
mnm_rss_shared_corZ+nullz	2	0.9973099	0	0.3665032	0.0332717	0.9667283
mnm_rss_naive_corZ+nullz	2	0.9990295	36	0.7231955	0.0264151	0.9735849
mnm_rss_ed_corZ+nullz	2	0.9991543	1	0.7119832	0.0296084	0.9703916
mnm_rss_ed_ddcan_corZ+nullz	2	0.9992644	6	0.7175893	0.0275404	0.9724596

UKB Bloodcells Mixture

tb = readRDS('output/ukb_rss_20210313/ukb_rss_20210313_cs/ukb_rss_cs_simuukb_bloodcells_mixture_glob.rds')
tb$method = rownames(tb)

Different residuals with Oracle prior

methods_resid = c('mnm_suff_oracle+oracle','mnm_rss_oracle+oracle', 
                  'mnm_rss_oracle+identity','mnm_rss_oracle+nullz','mnm_rss_oracle+corY')
rates_resid = tb %>% filter(method %in% methods_resid)
rates_resid$method = sapply(rates_resid$method, function(x) rename[[x]])
rates_resid$method = gsub(' Oracle prior', '', rates_resid$method)
rates_resid = rates_resid[match(c('mvSuSiE Oracle residual','mvSuSiE-RSS Oracle residual','mvSuSiE-RSS Identity residual',
                                  'mvSuSiE-RSS Y residual','mvSuSiE-RSS z residual'), rates_resid$method),]
rates_resid[,c('size', 'purity', 'overlap', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	power	fdr	coverage
mnm_suff_oracle+oracle	2	0.9981412	0.8156973	0.0243085	0.9756915
mnm_rss_oracle+oracle	2	0.9981412	0.8156973	0.0243085	0.9756915
mnm_rss_oracle+identity	2	0.9993815	0.7995795	0.4574417	0.5425583
mnm_rss_oracle+corY	2	0.9981411	0.8170988	0.0234506	0.9765494
mnm_rss_oracle+nullz	2	0.9979292	0.8156973	0.0218487	0.9781513

Different priors with Oracle residual

methods_prior_oracleresid = c('mnm_rss_oracle+oracle', 'mnm_rss_identity+oracle', 'mnm_rss_shared+oracle',
                              'mnm_rss_naive+oracle', 'mnm_rss_ed+oracle','mnm_rss_ed_ddcan+oracle')
rates_priors_oracleresid = tb %>% filter(method %in% methods_prior_oracleresid)
rates_priors_oracleresid$method = sapply(rates_priors_oracleresid$method, function(x) rename[[x]])
rates_priors_oracleresid$method = gsub(' Oracle residual', '', rates_priors_oracleresid$method)
rates_priors_oracleresid = rates_priors_oracleresid[match(c('mvSuSiE-RSS Oracle prior','mvSuSiE-RSS Fixed effects prior',
                                                            'mvSuSiE-RSS Random effects prior','mvSuSiE-RSS Default prior',
                                                            'mvSuSiE-RSS ED prior','mvSuSiE-RSS ED+default prior'), 
                                                          rates_priors_oracleresid$method),]
rates_priors_oracleresid[,c('size', 'purity', 'overlap', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	overlap	power	fdr	coverage
mnm_rss_oracle+oracle	2	0.9981412	0	0.8156973	0.0243085	0.9756915
mnm_rss_shared+oracle	16	0.9075882	0	0.0014015	0.3333333	0.6666667
mnm_rss_identity+oracle	2	0.9987416	0	0.7575333	0.0278777	0.9721223
mnm_rss_naive+oracle	2	0.9988085	3	0.7217940	0.0236967	0.9763033
mnm_rss_ed+oracle	2	0.9981934	0	0.8065872	0.0295110	0.9704890
mnm_rss_ed_ddcan+oracle	2	0.9982829	0	0.8058865	0.0287162	0.9712838

Different priors with estimated residual from Y

methods_prior_y = c('mnm_rss_oracle+corY', 'mnm_rss_identity_corY+corY',
                        'mnm_rss_shared_corY+corY', 'mnm_rss_naive_corY+corY',
                        "mnm_rss_ed_corY+corY","mnm_rss_ed_ddcan_corY+corY")
rates_priors_yresid = tb %>% filter(method %in% methods_prior_y)
rates_priors_yresid$method = sapply(rates_priors_yresid$method, function(x) rename[[x]])
rates_priors_yresid$method = gsub(' Y residual', '', rates_priors_yresid$method)
rates_priors_yresid = rates_priors_yresid[match(c('mvSuSiE-RSS Oracle prior','mvSuSiE-RSS Fixed effects prior',
                                                  'mvSuSiE-RSS Random effects prior','mvSuSiE-RSS Default prior',
                                                  'mvSuSiE-RSS ED prior','mvSuSiE-RSS ED+default prior'), 
                                                rates_priors_yresid$method),]
rates_priors_yresid[,c('size', 'purity', 'overlap', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	overlap	power	fdr	coverage
mnm_rss_oracle+corY	2	0.9981411	0	0.8170988	0.0234506	0.9765494
mnm_rss_shared_corY+corY	16	0.9075882	0	0.0014015	0.3333333	0.6666667
mnm_rss_identity_corY+corY	2	0.9987411	0	0.7582341	0.0261026	0.9738974
mnm_rss_naive_corY+corY	2	0.9988085	3	0.7224947	0.0227488	0.9772512
mnm_rss_ed_corY+corY	2	0.9981410	0	0.8072880	0.0294861	0.9705139
mnm_rss_ed_ddcan_corY+corY	2	0.9982829	0	0.8051857	0.0295608	0.9704392

Different priors with estimated residual from close to null z scores

methods_prior = c('mnm_rss_oracle+nullz', 'mnm_rss_identity_corZ+nullz', 
                  'mnm_rss_shared_corZ+nullz', 'mnm_rss_naive_corZ+nullz',
                  'mnm_rss_ed_corZ+nullz', 'mnm_rss_ed_ddcan_corZ+nullz')
rates_priors = tb %>% filter(method %in% methods_prior)
rates_priors$method = sapply(rates_priors$method, function(x) rename[[x]])
rates_priors$method = gsub(' Y residual', '', rates_priors$method)
rates_priors$method = gsub(' z residual', '', rates_priors$method)
rates_priors = rates_priors[match(c('mvSuSiE-RSS Oracle prior', 'mvSuSiE-RSS Random effects prior',
                                    'mvSuSiE-RSS Fixed effects prior','mvSuSiE-RSS Default prior',
                                    'mvSuSiE-RSS ED prior','mvSuSiE-RSS ED+default prior'), rates_priors$method),]
rates_priors[,c('size', 'purity', 'overlap', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	overlap	power	fdr	coverage
mnm_rss_oracle+nullz	2	0.9979292	0	0.8156973	0.0218487	0.9781513
mnm_rss_identity_corZ+nullz	2	0.9984231	0	0.7505256	0.0192308	0.9807692
mnm_rss_shared_corZ+nullz	22	0.8862773	0	0.0105116	0.1176471	0.8823529
mnm_rss_naive_corZ+nullz	2	0.9985588	2	0.7133847	0.0202117	0.9797883
mnm_rss_ed_corZ+nullz	2	0.9977725	0	0.8107919	0.0236287	0.9763713
mnm_rss_ed_ddcan_corZ+nullz	2	0.9979426	0	0.8079888	0.0237087	0.9762913

Trait specific comparison

We compare the CS and PIP for each SNP in each trait.

PIP calibration

Artificial Mixture

Different residuals with Oracle prior

Different priors with Oracle residual

Different priors with estimated residual from close to null z scores

UKB Bloodcells Mixture

Different residuals with Oracle prior

Different priors with Oracle residual

Different priors with estimated residual from close to null z scores

Power

Artificial Mixture

Different residuals with Oracle prior

Different priors with Oracle residual

Different priors with estimated residual from close to null z scores

UKB Bloodcells Mixture

Different residuals with Oracle prior

Different priors with Oracle residual

Different priors with estimated residual from close to null z scores

CS

Artificial Mixture

tb = readRDS('output/ukb_rss_20210313/ukb_rss_20210313_cs/ukb_rss_cs_simuartificial_mixture_ukb_cond.rds')
tb$method = rownames(tb)

Different residuals with Oracle prior

methods_resid = c('susie_suff+FALSE','susie_rss+FALSE','mnm_suff_oracle+oracle','mnm_rss_oracle+oracle', 
                  'mnm_rss_oracle+identity','mnm_rss_oracle+nullz','mnm_rss_oracle+corY')
rates_resid = tb %>% filter(method %in% methods_resid)
rates_resid$method = sapply(rates_resid$method, function(x) rename[[x]])
rates_resid$method = gsub(' Oracle prior', '', rates_resid$method)
rates_resid = rates_resid[match(c('SuSiE', 'SuSiE-RSS',
                                  'mvSuSiE Oracle residual','mvSuSiE-RSS Oracle residual','mvSuSiE-RSS Identity residual',
                                  'mvSuSiE-RSS Y residual','mvSuSiE-RSS z residual'), rates_resid$method),]
rates_resid[,c('size', 'purity', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	power	fdr	coverage
susie_suff+FALSE	4	0.9830877	0.4261258	0.0309749	0.9690251
susie_rss+FALSE	4	0.9830877	0.4279866	0.0314974	0.9685026
mnm_suff_oracle+oracle	1	1.0000000	0.7914403	0.0209024	0.9790976
mnm_rss_oracle+oracle	1	1.0000000	0.7914403	0.0209024	0.9790976
mnm_rss_oracle+identity	1	1.0000000	0.7914403	0.0209024	0.9790976
mnm_rss_oracle+corY	1	1.0000000	0.7912170	0.0209082	0.9790918
mnm_rss_oracle+nullz	1	1.0000000	0.7868999	0.0212924	0.9787076

Different priors with Oracle residual

methods_prior_oracleresid = c('susie_suff+FALSE','susie_rss+FALSE',
                              'mnm_rss_oracle+oracle', 'mnm_rss_identity+oracle', 'mnm_rss_shared+oracle',
                              'mnm_rss_naive+oracle', 'mnm_rss_ed+oracle','mnm_rss_ed_ddcan+oracle')
rates_priors_oracleresid = tb %>% filter(method %in% methods_prior_oracleresid)
rates_priors_oracleresid$method = sapply(rates_priors_oracleresid$method, function(x) rename[[x]])
rates_priors_oracleresid$method = gsub(' Oracle residual', '', rates_priors_oracleresid$method)
rates_priors_oracleresid = rates_priors_oracleresid[match(c('SuSiE', 'SuSiE-RSS',
                                                            'mvSuSiE-RSS Oracle prior','mvSuSiE-RSS Fixed effects prior',
                                                            'mvSuSiE-RSS Random effects prior','mvSuSiE-RSS Default prior',
                                                            'mvSuSiE-RSS ED prior','mvSuSiE-RSS ED+default prior'), 
                                                          rates_priors_oracleresid$method),]
rates_priors_oracleresid[,c('size', 'purity', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	power	fdr	coverage
susie_suff+FALSE	4	0.9830877	0.4261258	0.0309749	0.9690251
susie_rss+FALSE	4	0.9830877	0.4279866	0.0314974	0.9685026
mnm_rss_oracle+oracle	1	1.0000000	0.7914403	0.0209024	0.9790976
mnm_rss_shared+oracle	2	0.9976247	0.6875326	0.1334897	0.8665103
mnm_rss_identity+oracle	1	1.0000000	0.7234090	0.0747334	0.9252666
mnm_rss_naive+oracle	1	1.0000000	0.7426126	0.0377122	0.9622878
mnm_rss_ed+oracle	2	1.0000000	0.7847413	0.0496665	0.9503335
mnm_rss_ed_ddcan+oracle	1	1.0000000	0.7845925	0.0355901	0.9644099

Different priors with estimated residual from Y

methods_prior_y = c('susie_suff+FALSE','susie_rss+FALSE',
                    'mnm_rss_oracle+corY', 'mnm_rss_identity_corY+corY',
                    'mnm_rss_shared_corY+corY', 'mnm_rss_naive_corY+corY',
                    "mnm_rss_ed_corY+corY","mnm_rss_ed_ddcan_corY+corY")
rates_priors_yresid = tb %>% filter(method %in% methods_prior_y)
rates_priors_yresid$method = sapply(rates_priors_yresid$method, function(x) rename[[x]])
rates_priors_yresid$method = gsub(' Y residual', '', rates_priors_yresid$method)
rates_priors_yresid = rates_priors_yresid[match(c('SuSiE', 'SuSiE-RSS','mvSuSiE-RSS Oracle prior','mvSuSiE-RSS Fixed effects prior',
                                                  'mvSuSiE-RSS Random effects prior','mvSuSiE-RSS Default prior',
                                                  'mvSuSiE-RSS ED prior','mvSuSiE-RSS ED+default prior'), 
                                                rates_priors_yresid$method),]
rates_priors_yresid[,c('size', 'purity', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	power	fdr	coverage
susie_suff+FALSE	4	0.9830877	0.4261258	0.0309749	0.9690251
susie_rss+FALSE	4	0.9830877	0.4279866	0.0314974	0.9685026
mnm_rss_oracle+corY	1	1.0000000	0.7912170	0.0209082	0.9790918
mnm_rss_shared_corY+corY	2	0.9975845	0.6876814	0.1349251	0.8650749
mnm_rss_identity_corY+corY	1	1.0000000	0.7236323	0.0749762	0.9250238
mnm_rss_naive_corY+corY	1	1.0000000	0.7426126	0.0378050	0.9621950
mnm_rss_ed_corY+corY	1	1.0000000	0.7858578	0.0484859	0.9515141
mnm_rss_ed_ddcan_corY+corY	1	1.0000000	0.7846669	0.0355869	0.9644131

Different priors with estimated residual from close to null z scores

methods_prior = c('susie_suff+FALSE','susie_rss+FALSE',
                  'mnm_rss_oracle+nullz', 'mnm_rss_identity_corZ+nullz', 
                  'mnm_rss_shared_corZ+nullz', 'mnm_rss_naive_corZ+nullz',
                  'mnm_rss_ed_corZ+nullz', 'mnm_rss_ed_ddcan_corZ+nullz')
rates_priors = tb %>% filter(method %in% methods_prior)
rates_priors$method = sapply(rates_priors$method, function(x) rename[[x]])
rates_priors$method = gsub(' Y residual', '', rates_priors$method)
rates_priors$method = gsub(' z residual', '', rates_priors$method)
rates_priors = rates_priors[match(c('SuSiE', 'SuSiE-RSS','mvSuSiE-RSS Oracle prior', 'mvSuSiE-RSS Random effects prior',
                                    'mvSuSiE-RSS Fixed effects prior','mvSuSiE-RSS Default prior',
                                    'mvSuSiE-RSS ED prior','mvSuSiE-RSS ED+default prior'), rates_priors$method),]
rates_priors[,c('size', 'purity', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	power	fdr	coverage
susie_suff+FALSE	4	0.9830877	0.4261258	0.0309749	0.9690251
susie_rss+FALSE	4	0.9830877	0.4279866	0.0314974	0.9685026
mnm_rss_oracle+nullz	1	1.0000000	0.7868999	0.0212924	0.9787076
mnm_rss_identity_corZ+nullz	1	1.0000000	0.7217715	0.0792822	0.9207178
mnm_rss_shared_corZ+nullz	2	0.9973099	0.6817268	0.1535120	0.8464880
mnm_rss_naive_corZ+nullz	1	1.0000000	0.7379978	0.0422141	0.9577859
mnm_rss_ed_corZ+nullz	1	1.0000000	0.7816152	0.0501990	0.9498010
mnm_rss_ed_ddcan_corZ+nullz	1	1.0000000	0.7799032	0.0370370	0.9629630

UKB Bloodcells Mixture

tb = readRDS('output/ukb_rss_20210313/ukb_rss_20210313_cs/ukb_rss_cs_simuukb_bloodcells_mixture_cond.rds')
tb$method = rownames(tb)

Different residuals with Oracle prior

methods_resid = c('susie_suff+FALSE','susie_rss+FALSE','mnm_suff_oracle+oracle','mnm_rss_oracle+oracle', 
                  'mnm_rss_oracle+identity','mnm_rss_oracle+nullz','mnm_rss_oracle+corY')
rates_resid = tb %>% filter(method %in% methods_resid)
rates_resid$method = sapply(rates_resid$method, function(x) rename[[x]])
rates_resid$method = gsub(' Oracle prior', '', rates_resid$method)
rates_resid = rates_resid[match(c('SuSiE', 'SuSiE-RSS',
                                  'mvSuSiE Oracle residual','mvSuSiE-RSS Oracle residual','mvSuSiE-RSS Identity residual',
                                  'mvSuSiE-RSS Y residual','mvSuSiE-RSS z residual'), rates_resid$method),]
rates_resid[,c('size', 'purity', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	power	fdr	coverage
susie_suff+FALSE	5	0.9750062	0.1843465	0.0364011	0.9635989
susie_rss+FALSE	4	0.9756163	0.1843903	0.0368337	0.9631663
mnm_suff_oracle+oracle	2	0.9986987	0.5465137	0.0230956	0.9769044
mnm_rss_oracle+oracle	2	0.9986987	0.5465137	0.0230956	0.9769044
mnm_rss_oracle+identity	2	0.9995977	0.5678434	0.3566715	0.6433285
mnm_rss_oracle+corY	2	0.9986987	0.5477400	0.0221284	0.9778716
mnm_rss_oracle+nullz	2	0.9983670	0.5521198	0.0194462	0.9805538

Different priors with Oracle residual

methods_prior_oracleresid = c('susie_suff+FALSE','susie_rss+FALSE',
                              'mnm_rss_oracle+oracle', 'mnm_rss_identity+oracle', 'mnm_rss_shared+oracle',
                              'mnm_rss_naive+oracle', 'mnm_rss_ed+oracle','mnm_rss_ed_ddcan+oracle')
rates_priors_oracleresid = tb %>% filter(method %in% methods_prior_oracleresid)
rates_priors_oracleresid$method = sapply(rates_priors_oracleresid$method, function(x) rename[[x]])
rates_priors_oracleresid$method = gsub(' Oracle residual', '', rates_priors_oracleresid$method)
rates_priors_oracleresid = rates_priors_oracleresid[match(c('SuSiE', 'SuSiE-RSS',
                                                            'mvSuSiE-RSS Oracle prior','mvSuSiE-RSS Fixed effects prior',
                                                            'mvSuSiE-RSS Random effects prior','mvSuSiE-RSS Default prior',
                                                            'mvSuSiE-RSS ED prior','mvSuSiE-RSS ED+default prior'), 
                                                          rates_priors_oracleresid$method),]
rates_priors_oracleresid[,c('size', 'purity', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	power	fdr	coverage
susie_suff+FALSE	5	0.9750062	0.1843465	0.0364011	0.9635989
susie_rss+FALSE	4	0.9756163	0.1843903	0.0368337	0.9631663
mnm_rss_oracle+oracle	2	0.9986987	0.5465137	0.0230956	0.9769044
mnm_rss_shared+oracle	16	0.9075882	0.0014015	0.3333333	0.6666667
mnm_rss_identity+oracle	2	0.9993694	0.4535301	0.0206186	0.9793814
mnm_rss_naive+oracle	2	0.9995829	0.4331202	0.0196292	0.9803708
mnm_rss_ed+oracle	2	0.9991060	0.4830501	0.0254484	0.9745516
mnm_rss_ed_ddcan+oracle	2	0.9991060	0.4904958	0.0254960	0.9745040

Different priors with estimated residual from Y

methods_prior_y = c('susie_suff+FALSE','susie_rss+FALSE',
                    'mnm_rss_oracle+corY', 'mnm_rss_identity_corY+corY',
                    'mnm_rss_shared_corY+corY', 'mnm_rss_naive_corY+corY',
                    "mnm_rss_ed_corY+corY","mnm_rss_ed_ddcan_corY+corY")
rates_priors_yresid = tb %>% filter(method %in% methods_prior_y)
rates_priors_yresid$method = sapply(rates_priors_yresid$method, function(x) rename[[x]])
rates_priors_yresid$method = gsub(' Y residual', '', rates_priors_yresid$method)
rates_priors_yresid = rates_priors_yresid[match(c('SuSiE', 'SuSiE-RSS','mvSuSiE-RSS Oracle prior','mvSuSiE-RSS Fixed effects prior',
                                                  'mvSuSiE-RSS Random effects prior','mvSuSiE-RSS Default prior',
                                                  'mvSuSiE-RSS ED prior','mvSuSiE-RSS ED+default prior'), 
                                                rates_priors_yresid$method),]
rates_priors_yresid[,c('size', 'purity', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	power	fdr	coverage
susie_suff+FALSE	5	0.9750062	0.1843465	0.0364011	0.9635989
susie_rss+FALSE	4	0.9756163	0.1843903	0.0368337	0.9631663
mnm_rss_oracle+corY	2	0.9986987	0.5477400	0.0221284	0.9778716
mnm_rss_shared_corY+corY	16	0.9075882	0.0014015	0.3333333	0.6666667
mnm_rss_identity_corY+corY	2	0.9993655	0.4539681	0.0195800	0.9804200
mnm_rss_naive_corY+corY	2	0.9995751	0.4331202	0.0187537	0.9812463
mnm_rss_ed_corY+corY	2	0.9990928	0.4836195	0.0250750	0.9749250
mnm_rss_ed_ddcan_corY+corY	2	0.9991060	0.4898826	0.0256969	0.9743031

Different priors with estimated residual from close to null z scores

methods_prior = c('susie_suff+FALSE','susie_rss+FALSE',
                  'mnm_rss_oracle+nullz', 'mnm_rss_identity_corZ+nullz', 
                  'mnm_rss_shared_corZ+nullz', 'mnm_rss_naive_corZ+nullz',
                  'mnm_rss_ed_corZ+nullz', 'mnm_rss_ed_ddcan_corZ+nullz')
rates_priors = tb %>% filter(method %in% methods_prior)
rates_priors$method = sapply(rates_priors$method, function(x) rename[[x]])
rates_priors$method = gsub(' Y residual', '', rates_priors$method)
rates_priors$method = gsub(' z residual', '', rates_priors$method)
rates_priors = rates_priors[match(c('SuSiE', 'SuSiE-RSS','mvSuSiE-RSS Oracle prior', 'mvSuSiE-RSS Random effects prior',
                                    'mvSuSiE-RSS Fixed effects prior','mvSuSiE-RSS Default prior',
                                    'mvSuSiE-RSS ED prior','mvSuSiE-RSS ED+default prior'), rates_priors$method),]
rates_priors[,c('size', 'purity', 'power', 'fdr', 'coverage')] %>% kbl() %>% kable_styling()

	size	purity	power	fdr	coverage
susie_suff+FALSE	5	0.9750062	0.1843465	0.0364011	0.9635989
susie_rss+FALSE	4	0.9756163	0.1843903	0.0368337	0.9631663
mnm_rss_oracle+nullz	2	0.9983670	0.5521198	0.0194462	0.9805538
mnm_rss_identity_corZ+nullz	2	0.9991964	0.4513840	0.0183827	0.9816173
mnm_rss_shared_corZ+nullz	22	0.8862773	0.0105116	0.1176471	0.8823529
mnm_rss_naive_corZ+nullz	2	0.9993694	0.4296601	0.0187056	0.9812944
mnm_rss_ed_corZ+nullz	2	0.9988109	0.4861598	0.0208186	0.9791814
mnm_rss_ed_ddcan_corZ+nullz	2	0.9988085	0.4927733	0.0217372	0.9782628

sessionInfo()

R version 4.0.3 (2020-10-10)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Big Sur 10.16

Matrix products: default
BLAS:   /Library/Frameworks/R.framework/Versions/4.0/Resources/lib/libRblas.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/4.0/Resources/lib/libRlapack.dylib

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] kableExtra_1.3.4 dplyr_1.0.5      workflowr_1.6.2 

loaded via a namespace (and not attached):
 [1] Rcpp_1.0.6        highr_0.8         pillar_1.5.1      compiler_4.0.3   
 [5] later_1.1.0.1     git2r_0.28.0      tools_4.0.3       digest_0.6.27    
 [9] viridisLite_0.3.0 evaluate_0.14     lifecycle_1.0.0   tibble_3.1.0     
[13] pkgconfig_2.0.3   rlang_0.4.10      rstudioapi_0.13   DBI_1.1.1        
[17] yaml_2.2.1        xfun_0.22         xml2_1.3.2        httr_1.4.2       
[21] stringr_1.4.0     knitr_1.31        systemfonts_1.0.1 generics_0.1.0   
[25] fs_1.5.0          vctrs_0.3.7       webshot_0.5.2     rprojroot_2.0.2  
[29] tidyselect_1.1.0  svglite_2.0.0     glue_1.4.2        R6_2.5.0         
[33] fansi_0.4.2       rmarkdown_2.7     purrr_0.3.4       magrittr_2.0.1   
[37] whisker_0.4       scales_1.1.1      promises_1.2.0.1  ellipsis_0.3.1   
[41] htmltools_0.5.1.1 rvest_1.0.0       assertthat_0.2.1  colorspace_2.0-0 
[45] httpuv_1.5.5      utf8_1.2.1        stringi_1.5.3     munsell_0.5.0    
[49] crayon_1.4.1

mvSuSiE RSS ukb

Yuxin Zou

03/16/2021

Global comparison

PIP calibration

Power

CS

Trait specific comparison

PIP calibration

Power

CS