Last updated: 2025-06-04

Checks: 6 1

Knit directory: multigroup_ctwas_analysis/

This reproducible R Markdown analysis was created with workflowr (version 1.7.0). 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: uncommitted changes

The R Markdown is untracked by Git. To know which version of the R Markdown file created these results, you’ll want to first commit it to the Git repo. If you’re still working on the analysis, you can ignore this warning. When you’re finished, you can run wflow_publish to commit the R Markdown file and build the HTML.

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(20231112)

The command set.seed(20231112) 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: 4f9814a

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 4f9814a. 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:    .Rhistory
    Ignored:    cv/
    Ignored:    figures/lz/ADHD-ieu-a-1183/
    Ignored:    figures/lz/ASD-ieu-a-1185/
    Ignored:    figures/lz/IBD-ebi-a-GCST004131/single/
    Ignored:    figures/lz/MDD-ieu-b-102/single/
    Ignored:    figures/lz/PD-ieu-b-7/single/
    Ignored:    figures/lz/RA-panukb/single/

Untracked files:
    Untracked:  analysis/realdata_final_combine_qtls_3qtls.Rmd
    Untracked:  analysis/realdata_final_multigroup_validation_silver_3qtls.Rmd
    Untracked:  analysis/realdata_final_multigroup_validation_weighted_pwy_3qtls.Rmd

Unstaged changes:
    Modified:   analysis/index.Rmd
    Deleted:    analysis/realdata_final_genefunctions_3qtls.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.

There are no past versions. Publish this analysis with wflow_publish() to start tracking its development.

library(kableExtra)
library(tidyverse)

-- Attaching packages --------------------------------------- tidyverse 1.3.1 --

v ggplot2 3.5.1     v purrr   1.0.2
v tibble  3.2.1     v dplyr   1.1.4
v tidyr   1.3.0     v stringr 1.5.1
v readr   2.1.2     v forcats 0.5.1

-- Conflicts ------------------------------------------ tidyverse_conflicts() --
x dplyr::filter()     masks stats::filter()
x dplyr::group_rows() masks kableExtra::group_rows()
x dplyr::lag()        masks stats::lag()

Introduction

We validate genes with susie pip > 0.8 here: https://sq-96.github.io/multigroup_ctwas_analysis/realdata_final_multigroup_summary.html

The basic idea is:

Some biological pathways are related to the traits. Genes within these pathways are more likely to be associated with these traits. Our approach involves aggregating these genes into a collective group. This allows us to assess whether the genes identified by cTWAS are overrepresented in this group.

However, the presence of common genes across multiple pathways presents a challenge to this straightforward aggregation approach. To address this, we propose weighting the pathways, assigning a unique score to each gene. By selecting genes that meet a specific score threshold, we can form a more refined group. We can then evaluate the enrichment of cTWAS-identified genes within this selectively grouped set.

Model

The model is \(y=X*w\)

y is an n-dimensional vector representing gene-trait associations (n = number of genes), which can be:

z-scores computed by MAGMA
a binary vector indicating gene-trait relationships.

We tried different settings for the binary vector:

genes with FDR 0.05 as per MAGMA are marked 1.
Genes ranked in the top 500 by MAGMA p-values or meeting the FDR 0.05 threshold were labeled as 1.
Genes ranked in the top 1000 by MAGMA p-values or meeting the FDR 0.05 threshold were labeled as 1.

X is an n×m matrix (m = number of pathways) indicating gene membership in specific pathways.

We fitted this model using different models.

If y is a z-score vector, it can be fitted using

- XGBoost: regression with squared loss

If y is a binarized vector, the model can be fitted using

- XGBoost: logistic regression for binary classification, output probability

Benchmarks

The model fitting results in pathway weights, from which we predict gene labels \(\hat{y}\). We then categorize genes based on these new labels.

For z-score model, we compute the p-values from the new labels(z-scores), then compute FDR. Then we tested different cutoffs for gene selection. The cutoffs are: 0.05,0.1,0.2
For binarized model. Genes with labels > 0.5/0.6/0.7/0.8 are considered benchmarks.

Testing genes

Genes from ctwas results are divided into different groups based on their SuSiE PIPs:

high (>0.8)
moderate (0.8 > PIP > 0.5)
low (<0.5)

Fisher exact test

We assess whether high-PIP genes are more enriched in our benchmarks than other groups using Fisher exact tests.

The testing matrix is:

fisher_matrix <- matrix(c("n1","n2","n3","n4"),nrow = 2,ncol = 2)
rownames(fisher_matrix) <- c("#included","#notincluded")
colnames(fisher_matrix) <- c("pip08","other group")

print(fisher_matrix)

             pip08 other group
#included    "n1"  "n3"       
#notincluded "n2"  "n4"

Pathways

The pathways are from Go Biological Process (gobp), Go Molecular Function (gomf), Go Cellular Component (gocc) and KEGG.

Results

GOBP

Binary, not truncated

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_prob0.5_gobp.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.5, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.5, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	64	73	5394	5321	73	0.00059	0.00039	1.00000
IBD-ebi-a-GCST004131	115	28	3749	3699	50	0.00205	0.00175	0.44778
aFib-ebi-a-GCST006414	36	53	3540	3489	51	0.00030	0.00022	0.67849
SBP-ukb-a-360	126	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	70	17	2855	2822	31	0.23603	0.22910	1.00000
Height-panukb	15148	208	13763	13584	179	0.00740	0.00737	0.87110
HTN-panukb	108	42	5465	5400	65	0.03689	0.03266	1.00000
PLT-panukb	15219	119	10616	10471	145	0.33397	0.33579	0.04387
RA-panukb	27	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	10362	122	10154	9976	178	0.00056	0.00055	0.69153
ATH_gtexukb	70	21	3276	3227	49	0.03517	0.03336	0.57803
BMI-panukb	13394	53	9215	9114	101	0.75242	0.75246	0.39287
HB-panukb	1112	90	8649	8522	127	0.02761	0.02634	0.84386
T2D-panukb	17	14	2460	2423	37	0.06597	0.06695	0.27451
SCZ-ieu-b-5102	190	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	43	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	6	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	2	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	10	4	726	717	9	0.00047	0.00048	0.07692
NS-ukb-a-230	48	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	18	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	13703	243	11369	11103	266	0.00016	0.00012	0.38708

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_prob0.6_gobp.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.6, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.6, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	24	73	5394	5321	73	0.23611	0.20704	0.61977
IBD-ebi-a-GCST004131	76	28	3749	3699	50	0.00055	0.00046	0.24286
aFib-ebi-a-GCST006414	20	53	3540	3489	51	0.00178	0.00186	0.24300
SBP-ukb-a-360	53	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	35	17	2855	2822	31	1.00000	1.00000	0.53280
Height-panukb	15028	208	13763	13584	179	0.00583	0.00454	0.87110
HTN-panukb	44	42	5465	5400	65	1.00000	1.00000	0.51860
PLT-panukb	6897	119	10616	10471	145	0.00095	0.00070	0.08313
RA-panukb	23	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	1491	122	10154	9976	178	0.00001	0.00000	0.32109
ATH_gtexukb	36	21	3276	3227	49	0.15904	0.16125	0.30000
BMI-panukb	1194	53	9215	9114	101	0.11045	0.10709	1.00000
HB-panukb	233	90	8649	8522	127	0.08272	0.08319	0.23524
T2D-panukb	7	14	2460	2423	37	1.00000	1.00000	1.00000
SCZ-ieu-b-5102	97	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	9	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	3	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	2	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	6	4	726	717	9	0.02717	0.02751	0.30769
NS-ukb-a-230	21	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	7	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	1617	243	11369	11103	266	0.00000	0.00000	0.25095

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_prob0.7_gobp.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.7, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.7, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	10	73	5394	5321	73	0.10202	0.07854	1.00000
IBD-ebi-a-GCST004131	46	28	3749	3699	50	0.02292	0.01950	1.00000
aFib-ebi-a-GCST006414	7	53	3540	3489	51	0.00428	0.00440	0.49533
SBP-ukb-a-360	18	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	24	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	14703	208	13763	13584	179	0.00431	0.00337	1.00000
HTN-panukb	15	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	418	119	10616	10471	145	0.77556	0.77460	0.19290
RA-panukb	18	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	283	122	10154	9976	178	0.00360	0.00326	0.41933
ATH_gtexukb	27	21	3276	3227	49	0.13153	0.13339	0.30000
BMI-panukb	39	53	9215	9114	101	0.16312	0.16477	0.34416
HB-panukb	41	90	8649	8522	127	0.04260	0.04374	0.17089
T2D-panukb	3	14	2460	2423	37	1.00000	1.00000	1.00000
SCZ-ieu-b-5102	51	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	0	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	0	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	1	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	4	4	726	717	9	0.02178	0.02205	0.30769
NS-ukb-a-230	9	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	1	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	108	243	11369	11103	266	0.10926	0.09895	1.00000

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_prob0.8_gobp.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.8, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.8, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	3	73	5394	5321	73	1.00000	1.00000	1.00000
IBD-ebi-a-GCST004131	19	28	3749	3699	50	0.09909	0.09353	1.00000
aFib-ebi-a-GCST006414	3	53	3540	3489	51	0.04362	0.04423	1.00000
SBP-ukb-a-360	4	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	9	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	7348	208	13763	13584	179	0.00228	0.00224	0.91891
HTN-panukb	1	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	26	119	10616	10471	145	1.00000	1.00000	1.00000
RA-panukb	14	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	36	122	10154	9976	178	0.30149	0.28880	1.00000
ATH_gtexukb	5	21	3276	3227	49	1.00000	1.00000	1.00000
BMI-panukb	0	53	9215	9114	101	1.00000	1.00000	1.00000
HB-panukb	4	90	8649	8522	127	1.00000	1.00000	1.00000
T2D-panukb	2	14	2460	2423	37	1.00000	1.00000	1.00000
SCZ-ieu-b-5102	4	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	0	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	0	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	1	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	1	4	726	717	9	1.00000	1.00000	1.00000
NS-ukb-a-230	1	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	0	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	0	243	11369	11103	266	1.00000	1.00000	1.00000

Binary, truncated top 500 genes

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc500_prob0.5_gobp.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.5, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.5, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	33	73	5394	5321	73	0.00000	0.00000	0.74505
IBD-ebi-a-GCST004131	41	28	3749	3699	50	0.00010	0.00010	0.01435
aFib-ebi-a-GCST006414	23	53	3540	3489	51	0.02521	0.02306	1.00000
SBP-ukb-a-360	7	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	48	17	2855	2822	31	1.00000	1.00000	0.53280
Height-panukb	35	208	13763	13584	179	1.00000	1.00000	0.46253
HTN-panukb	10	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	22	119	10616	10471	145	0.20000	0.19337	1.00000
RA-panukb	25	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	25	122	10154	9976	178	0.25839	0.25328	1.00000
ATH_gtexukb	54	21	3276	3227	49	0.00233	0.00212	0.15514
BMI-panukb	11	53	9215	9114	101	0.05033	0.05087	0.34416
HB-panukb	20	90	8649	8522	127	1.00000	1.00000	1.00000
T2D-panukb	20	14	2460	2423	37	0.08182	0.07769	0.47765
SCZ-ieu-b-5102	25	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	19	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	6	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	2	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	9	4	726	717	9	0.03254	0.03294	0.30769
NS-ukb-a-230	17	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	26	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	43	243	11369	11103	266	0.05391	0.05041	0.67332

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc500_prob0.6_gobp.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.6, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.6, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	24	73	5394	5321	73	0.00021	0.00010	1.00000
IBD-ebi-a-GCST004131	31	28	3749	3699	50	0.00075	0.00078	0.04306
aFib-ebi-a-GCST006414	11	53	3540	3489	51	0.09888	0.10024	1.00000
SBP-ukb-a-360	3	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	36	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	22	208	13763	13584	179	1.00000	1.00000	1.00000
HTN-panukb	7	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	14	119	10616	10471	145	1.00000	1.00000	1.00000
RA-panukb	23	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	19	122	10154	9976	178	0.20318	0.20638	0.40667
ATH_gtexukb	39	21	3276	3227	49	0.18571	0.18292	0.51304
BMI-panukb	6	53	9215	9114	101	1.00000	1.00000	1.00000
HB-panukb	14	90	8649	8522	127	1.00000	1.00000	1.00000
T2D-panukb	15	14	2460	2423	37	0.07129	0.06695	0.47765
SCZ-ieu-b-5102	22	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	3	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	6	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	2	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	5	4	726	717	9	0.02717	0.02751	0.30769
NS-ukb-a-230	11	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	14	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	27	243	11369	11103	266	0.43541	0.43080	1.00000

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc500_prob0.7_gobp.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.7, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.7, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	11	73	5394	5321	73	0.00026	0.00013	1.00000
IBD-ebi-a-GCST004131	17	28	3749	3699	50	0.09909	0.10036	0.35897
aFib-ebi-a-GCST006414	6	53	3540	3489	51	0.05773	0.05855	1.00000
SBP-ukb-a-360	0	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	28	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	12	208	13763	13584	179	1.00000	1.00000	1.00000
HTN-panukb	2	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	8	119	10616	10471	145	1.00000	1.00000	1.00000
RA-panukb	18	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	14	122	10154	9976	178	1.00000	1.00000	1.00000
ATH_gtexukb	29	21	3276	3227	49	0.14814	0.15021	0.30000
BMI-panukb	2	53	9215	9114	101	1.00000	1.00000	1.00000
HB-panukb	11	90	8649	8522	127	1.00000	1.00000	1.00000
T2D-panukb	8	14	2460	2423	37	0.04445	0.03957	0.47765
SCZ-ieu-b-5102	8	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	1	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	1	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	1	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	3	4	726	717	9	0.01637	0.01657	0.30769
NS-ukb-a-230	5	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	6	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	15	243	11369	11103	266	1.00000	1.00000	1.00000

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc500_prob0.8_gobp.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.8, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.8, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	9	73	5394	5321	73	0.00012	0.00008	0.61977
IBD-ebi-a-GCST004131	5	28	3749	3699	50	0.02934	0.02973	0.35897
aFib-ebi-a-GCST006414	1	53	3540	3489	51	1.00000	1.00000	1.00000
SBP-ukb-a-360	0	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	11	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	5	208	13763	13584	179	1.00000	1.00000	1.00000
HTN-panukb	0	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	1	119	10616	10471	145	1.00000	1.00000	1.00000
RA-panukb	14	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	7	122	10154	9976	178	1.00000	1.00000	1.00000
ATH_gtexukb	13	21	3276	3227	49	1.00000	1.00000	1.00000
BMI-panukb	0	53	9215	9114	101	1.00000	1.00000	1.00000
HB-panukb	4	90	8649	8522	127	1.00000	1.00000	1.00000
T2D-panukb	1	14	2460	2423	37	1.00000	1.00000	1.00000
SCZ-ieu-b-5102	7	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	0	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	0	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	0	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	0	4	726	717	9	1.00000	1.00000	1.00000
NS-ukb-a-230	1	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	3	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	4	243	11369	11103	266	1.00000	1.00000	1.00000

Binary, truncated top 1000 genes

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc1000_prob0.5_gobp.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.5, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.5, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	47	73	5394	5321	73	0.00244	0.00178	1.00000
IBD-ebi-a-GCST004131	73	28	3749	3699	50	0.00073	0.00072	0.05315
aFib-ebi-a-GCST006414	23	53	3540	3489	51	0.00078	0.00063	0.61790
SBP-ukb-a-360	20	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	50	17	2855	2822	31	1.00000	1.00000	0.54296
Height-panukb	47	208	13763	13584	179	0.13926	0.13700	1.00000
HTN-panukb	16	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	37	119	10616	10471	145	0.05147	0.04711	1.00000
RA-panukb	27	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	27	122	10154	9976	178	0.26722	0.27125	0.40667
ATH_gtexukb	56	21	3276	3227	49	0.03245	0.03066	0.57803
BMI-panukb	14	53	9215	9114	101	1.00000	1.00000	1.00000
HB-panukb	41	90	8649	8522	127	0.23627	0.23126	1.00000
T2D-panukb	9	14	2460	2423	37	1.00000	1.00000	1.00000
SCZ-ieu-b-5102	42	15	3547	3511	36	0.08504	0.08194	0.50588
BIP-ieu-b-5110	45	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	7	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	3	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	9	4	726	717	9	0.03254	0.03294	0.30769
NS-ukb-a-230	40	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	19	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	52	243	11369	11103	266	0.00449	0.00350	0.74270

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc500_prob0.6_gobp.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.6, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.6, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	24	73	5394	5321	73	0.00021	0.00010	1.00000
IBD-ebi-a-GCST004131	31	28	3749	3699	50	0.00075	0.00078	0.04306
aFib-ebi-a-GCST006414	11	53	3540	3489	51	0.09888	0.10024	1.00000
SBP-ukb-a-360	3	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	36	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	22	208	13763	13584	179	1.00000	1.00000	1.00000
HTN-panukb	7	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	14	119	10616	10471	145	1.00000	1.00000	1.00000
RA-panukb	23	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	19	122	10154	9976	178	0.20318	0.20638	0.40667
ATH_gtexukb	39	21	3276	3227	49	0.18571	0.18292	0.51304
BMI-panukb	6	53	9215	9114	101	1.00000	1.00000	1.00000
HB-panukb	14	90	8649	8522	127	1.00000	1.00000	1.00000
T2D-panukb	15	14	2460	2423	37	0.07129	0.06695	0.47765
SCZ-ieu-b-5102	22	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	3	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	6	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	2	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	5	4	726	717	9	0.02717	0.02751	0.30769
NS-ukb-a-230	11	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	14	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	27	243	11369	11103	266	0.43541	0.43080	1.00000

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc1000_prob0.7_gobp.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.7, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.7, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	22	73	5394	5321	73	0.00221	0.00168	1.00000
IBD-ebi-a-GCST004131	25	28	3749	3699	50	0.01117	0.01146	0.12587
aFib-ebi-a-GCST006414	4	53	3540	3489	51	0.02929	0.02971	1.00000
SBP-ukb-a-360	5	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	22	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	14	208	13763	13584	179	0.16479	0.15398	1.00000
HTN-panukb	0	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	9	119	10616	10471	145	1.00000	1.00000	1.00000
RA-panukb	20	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	11	122	10154	9976	178	0.12317	0.12521	0.40667
ATH_gtexukb	27	21	3276	3227	49	0.13710	0.13903	0.30000
BMI-panukb	4	53	9215	9114	101	1.00000	1.00000	1.00000
HB-panukb	4	90	8649	8522	127	1.00000	1.00000	1.00000
T2D-panukb	1	14	2460	2423	37	1.00000	1.00000	1.00000
SCZ-ieu-b-5102	7	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	0	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	1	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	1	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	3	4	726	717	9	0.01637	0.01657	0.30769
NS-ukb-a-230	6	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	1	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	10	243	11369	11103	266	0.19069	0.17710	1.00000

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc1000_prob0.8_gobp.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.8, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.8, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	8	73	5394	5321	73	0.08986	0.07854	1.00000
IBD-ebi-a-GCST004131	15	28	3749	3699	50	0.09234	0.09353	0.35897
aFib-ebi-a-GCST006414	1	53	3540	3489	51	0.01475	0.01496	1.00000
SBP-ukb-a-360	1	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	8	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	5	208	13763	13584	179	0.07227	0.07318	1.00000
HTN-panukb	0	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	3	119	10616	10471	145	1.00000	1.00000	1.00000
RA-panukb	14	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	1	122	10154	9976	178	1.00000	1.00000	1.00000
ATH_gtexukb	6	21	3276	3227	49	1.00000	1.00000	1.00000
BMI-panukb	1	53	9215	9114	101	1.00000	1.00000	1.00000
HB-panukb	0	90	8649	8522	127	1.00000	1.00000	1.00000
T2D-panukb	0	14	2460	2423	37	1.00000	1.00000	1.00000
SCZ-ieu-b-5102	0	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	0	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	0	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	0	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	0	4	726	717	9	1.00000	1.00000	1.00000
NS-ukb-a-230	1	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	0	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	3	243	11369	11103	266	0.06148	0.04238	1.00000

Zscores

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_zscore/results_zfdr0.05_gobp.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff FDR 0.05, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff FDR 0.05, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	75	73	5394	5321	73	0.00243	0.00199	1.00000
IBD-ebi-a-GCST004131	40	28	3749	3699	50	0.00000	0.00001	0.00466
aFib-ebi-a-GCST006414	7	53	3540	3489	51	0.00428	0.00317	1.00000
SBP-ukb-a-360	6	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	27	17	2855	2822	31	0.14873	0.15033	0.35417
Height-panukb	15515	208	13763	13584	179	0.00933	0.00934	0.87110
HTN-panukb	9	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	15484	119	10616	10471	145	0.45810	0.45922	0.11976
RA-panukb	15	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	14551	122	10154	9976	178	0.08884	0.08885	0.86845
ATH_gtexukb	37	21	3276	3227	49	0.02026	0.01972	0.21228
BMI-panukb	14562	53	9215	9114	101	0.29857	0.29850	1.00000
HB-panukb	169	90	8649	8522	127	0.38835	0.36547	0.20077
T2D-panukb	4	14	2460	2423	37	1.00000	1.00000	1.00000
SCZ-ieu-b-5102	25	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	0	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	0	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	0	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	2	4	726	717	9	0.01094	0.01107	0.30769
NS-ukb-a-230	2	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	0	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	14788	243	11369	11103	266	0.00247	0.00194	0.68161

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_zscore/results_zfdr0.1_gobp.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff FDR 0.1, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff FDR 0.1, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	78	73	5394	5321	73	0.00042	0.00033	0.74505
IBD-ebi-a-GCST004131	45	28	3749	3699	50	0.00001	0.00001	0.00466
aFib-ebi-a-GCST006414	11	53	3540	3489	51	0.00000	0.00000	0.20550
SBP-ukb-a-360	7	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	27	17	2855	2822	31	0.14873	0.15033	0.35417
Height-panukb	15679	208	13763	13584	179	0.01178	0.01185	0.87110
HTN-panukb	16	42	5465	5400	65	0.10175	0.09591	1.00000
PLT-panukb	15534	119	10616	10471	145	0.38775	0.45681	0.11976
RA-panukb	17	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	15063	122	10154	9976	178	0.15725	0.15738	0.73770
ATH_gtexukb	43	21	3276	3227	49	0.02251	0.02198	0.21228
BMI-panukb	15131	53	9215	9114	101	0.47304	0.47293	1.00000
HB-panukb	14479	90	8649	8522	127	0.00077	0.00078	0.08305
T2D-panukb	4	14	2460	2423	37	1.00000	1.00000	1.00000
SCZ-ieu-b-5102	32	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	0	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	0	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	0	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	3	4	726	717	9	0.01637	0.01657	0.30769
NS-ukb-a-230	2	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	0	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	15423	243	11369	11103	266	0.01223	0.00977	0.68161

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_zscore/results_zfdr0.2_gobp.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff FDR 0.2, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff FDR 0.2, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	92	73	5394	5321	73	0.00088	0.00073	0.74505
IBD-ebi-a-GCST004131	47	28	3749	3699	50	0.00000	0.00000	0.00147
aFib-ebi-a-GCST006414	14	53	3540	3489	51	0.00000	0.00000	0.20550
SBP-ukb-a-360	7	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	33	17	2855	2822	31	0.16891	0.16058	1.00000
Height-panukb	15751	208	13763	13584	179	0.01482	0.01179	0.87110
HTN-panukb	19	42	5465	5400	65	0.11544	0.10985	1.00000
PLT-panukb	15578	119	10616	10471	145	0.38265	0.38415	0.08045
RA-panukb	17	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	15704	122	10154	9976	178	0.27785	0.23099	0.73851
ATH_gtexukb	44	21	3276	3227	49	0.02251	0.02198	0.21228
BMI-panukb	15275	53	9215	9114	101	0.58372	0.58390	1.00000
HB-panukb	15520	90	8649	8522	127	0.00337	0.00336	0.08305
T2D-panukb	6	14	2460	2423	37	1.00000	1.00000	1.00000
SCZ-ieu-b-5102	57	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	0	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	0	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	0	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	3	4	726	717	9	0.01637	0.01657	0.30769
NS-ukb-a-230	2	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	0	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	15680	243	11369	11103	266	0.02287	0.01847	0.68161

GOBP + GOCC + GOMF + KEGG jointly

Binary, not truncated

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_prob0.5_all.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.5, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.5, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	89	73	5394	5321	73	0.00014	0.00010	0.74505
IBD-ebi-a-GCST004131	111	28	3749	3699	50	0.00000	0.00000	0.00266
aFib-ebi-a-GCST006414	34	53	3540	3489	51	0.00030	0.00027	0.36316
SBP-ukb-a-360	112	35	4909	4858	51	0.00732	0.00682	0.39306
T1D-GCST90014023	122	17	2855	2822	31	1.00000	1.00000	0.54296
Height-panukb	16807	208	13763	13584	179	0.00043	0.00043	0.32667
HTN-panukb	199	42	5465	5400	65	1.00000	1.00000	0.15414
PLT-panukb	16707	119	10616	10471	145	0.07674	0.07670	0.48462
RA-panukb	30	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	11994	122	10154	9976	178	0.01093	0.01086	1.00000
ATH_gtexukb	120	21	3276	3227	49	0.04533	0.04657	0.08696
BMI-panukb	14025	53	9215	9114	101	0.41054	0.41050	1.00000
HB-panukb	1462	90	8649	8522	127	0.07592	0.07363	1.00000
T2D-panukb	29	14	2460	2423	37	0.10257	0.10405	0.27451
SCZ-ieu-b-5102	303	15	3547	3511	36	0.36496	0.36237	1.00000
BIP-ieu-b-5110	104	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	6	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	1	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	11	4	726	717	9	0.00080	0.00082	0.07692
NS-ukb-a-230	51	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	35	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	14554	243	11369	11103	266	0.00006	0.00005	1.00000

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_prob0.6_all.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.6, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.6, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	50	73	5394	5321	73	0.00313	0.00256	1.00000
IBD-ebi-a-GCST004131	71	28	3749	3699	50	0.00002	0.00002	0.02060
aFib-ebi-a-GCST006414	12	53	3540	3489	51	0.00719	0.00581	1.00000
SBP-ukb-a-360	41	35	4909	4858	51	0.00772	0.00788	0.16279
T1D-GCST90014023	81	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	16646	208	13763	13584	179	0.00059	0.00042	0.47319
HTN-panukb	128	42	5465	5400	65	1.00000	1.00000	0.27791
PLT-panukb	7535	119	10616	10471	145	0.00038	0.00027	0.53086
RA-panukb	26	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	1861	122	10154	9976	178	0.00000	0.00000	0.21649
ATH_gtexukb	64	21	3276	3227	49	0.19616	0.19882	0.30000
BMI-panukb	1789	53	9215	9114	101	0.17654	0.17453	0.82876
HB-panukb	319	90	8649	8522	127	0.04774	0.04665	0.49425
T2D-panukb	15	14	2460	2423	37	0.06063	0.06153	0.27451
SCZ-ieu-b-5102	185	15	3547	3511	36	0.22534	0.22065	1.00000
BIP-ieu-b-5110	68	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	4	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	1	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	7	4	726	717	9	0.00034	0.00034	0.07692
NS-ukb-a-230	24	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	25	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	2689	243	11369	11103	266	0.00000	0.00000	0.34379

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_prob0.7_all.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.7, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.7, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	20	73	5394	5321	73	0.02344	0.01918	1.00000
IBD-ebi-a-GCST004131	38	28	3749	3699	50	0.00117	0.00110	0.12919
aFib-ebi-a-GCST006414	7	53	3540	3489	51	0.07165	0.05855	1.00000
SBP-ukb-a-360	17	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	56	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	16036	208	13763	13584	179	0.00050	0.00050	0.52703
HTN-panukb	76	42	5465	5400	65	1.00000	1.00000	0.51860
PLT-panukb	720	119	10616	10471	145	0.07638	0.07481	0.82302
RA-panukb	24	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	313	122	10154	9976	178	0.03521	0.03415	0.36371
ATH_gtexukb	43	21	3276	3227	49	0.15361	0.15575	0.30000
BMI-panukb	248	53	9215	9114	101	0.00524	0.00499	0.23375
HB-panukb	91	90	8649	8522	127	0.34650	0.35062	0.41475
T2D-panukb	10	14	2460	2423	37	0.03899	0.03957	0.27451
SCZ-ieu-b-5102	104	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	16	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	2	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	1	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	4	4	726	717	9	0.01637	0.01657	0.30769
NS-ukb-a-230	10	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	16	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	501	243	11369	11103	266	0.00168	0.00088	0.72005

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_prob0.8_all.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.8, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.8, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	6	73	5394	5321	73	1.00000	1.00000	0.49655
IBD-ebi-a-GCST004131	22	28	3749	3699	50	0.12562	0.12057	1.00000
aFib-ebi-a-GCST006414	1	53	3540	3489	51	1.00000	1.00000	1.00000
SBP-ukb-a-360	4	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	24	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	7198	208	13763	13584	179	0.08952	0.08877	0.26242
HTN-panukb	15	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	77	119	10616	10471	145	1.00000	1.00000	0.25445
RA-panukb	21	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	71	122	10154	9976	178	0.05813	0.05995	0.16457
ATH_gtexukb	21	21	3276	3227	49	1.00000	1.00000	1.00000
BMI-panukb	10	53	9215	9114	101	1.00000	1.00000	1.00000
HB-panukb	44	90	8649	8522	127	1.00000	1.00000	1.00000
T2D-panukb	7	14	2460	2423	37	0.03351	0.03401	0.27451
SCZ-ieu-b-5102	21	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	5	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	0	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	1	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	0	4	726	717	9	1.00000	1.00000	1.00000
NS-ukb-a-230	5	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	1	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	123	243	11369	11103	266	0.35855	0.35371	1.00000

Binary, truncated top 500 genes

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc500_prob0.5_all.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.5, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.5, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	36	73	5394	5321	73	0.00100	0.00074	1.00000
IBD-ebi-a-GCST004131	56	28	3749	3699	50	0.00038	0.00036	0.05315
aFib-ebi-a-GCST006414	26	53	3540	3489	51	0.00025	0.00022	0.36316
SBP-ukb-a-360	9	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	63	17	2855	2822	31	0.21739	0.20525	1.00000
Height-panukb	46	208	13763	13584	179	0.12915	0.12680	1.00000
HTN-panukb	9	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	30	119	10616	10471	145	0.27653	0.27154	1.00000
RA-panukb	29	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	21	122	10154	9976	178	0.22202	0.21599	1.00000
ATH_gtexukb	68	21	3276	3227	49	0.03656	0.03336	0.63237
BMI-panukb	9	53	9215	9114	101	1.00000	1.00000	1.00000
HB-panukb	39	90	8649	8522	127	1.00000	1.00000	1.00000
T2D-panukb	19	14	2460	2423	37	0.09742	0.09359	0.47765
SCZ-ieu-b-5102	68	15	3547	3511	36	0.10426	0.09756	1.00000
BIP-ieu-b-5110	26	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	6	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	4	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	11	4	726	717	9	0.00080	0.00082	0.07692
NS-ukb-a-230	13	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	30	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	38	243	11369	11103	266	0.04463	0.04419	0.35238

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc500_prob0.6_all.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.6, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.6, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	27	73	5394	5321	73	0.00035	0.00023	1.00000
IBD-ebi-a-GCST004131	45	28	3749	3699	50	0.00279	0.00290	0.04306
aFib-ebi-a-GCST006414	13	53	3540	3489	51	0.00891	0.00740	1.00000
SBP-ukb-a-360	2	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	44	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	31	208	13763	13584	179	0.34322	0.33680	1.00000
HTN-panukb	6	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	21	119	10616	10471	145	1.00000	1.00000	1.00000
RA-panukb	25	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	15	122	10154	9976	178	0.16412	0.16678	0.40667
ATH_gtexukb	44	21	3276	3227	49	0.01918	0.01757	0.57803
BMI-panukb	3	53	9215	9114	101	1.00000	1.00000	1.00000
HB-panukb	14	90	8649	8522	127	1.00000	1.00000	1.00000
T2D-panukb	12	14	2460	2423	37	0.06063	0.05609	0.47765
SCZ-ieu-b-5102	37	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	1	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	4	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	1	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	6	4	726	717	9	0.02717	0.02751	0.30769
NS-ukb-a-230	8	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	22	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	27	243	11369	11103	266	0.43541	0.44302	0.47741

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc500_prob0.7_all.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.7, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.7, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	21	73	5394	5321	73	0.00012	0.00008	0.68090
IBD-ebi-a-GCST004131	25	28	3749	3699	50	0.00051	0.00053	0.04306
aFib-ebi-a-GCST006414	6	53	3540	3489	51	0.00208	0.00213	0.49533
SBP-ukb-a-360	1	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	30	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	17	208	13763	13584	179	0.21347	0.20392	1.00000
HTN-panukb	4	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	13	119	10616	10471	145	1.00000	1.00000	1.00000
RA-panukb	23	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	13	122	10154	9976	178	1.00000	1.00000	1.00000
ATH_gtexukb	31	21	3276	3227	49	0.01322	0.01360	0.08696
BMI-panukb	1	53	9215	9114	101	1.00000	1.00000	1.00000
HB-panukb	12	90	8649	8522	127	1.00000	1.00000	1.00000
T2D-panukb	11	14	2460	2423	37	0.05527	0.05061	0.47765
SCZ-ieu-b-5102	13	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	0	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	2	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	1	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	3	4	726	717	9	0.01637	0.01657	0.30769
NS-ukb-a-230	1	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	15	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	20	243	11369	11103	266	0.34513	0.35168	0.47741

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc500_prob0.8_all.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.8, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.8, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	13	73	5394	5321	73	0.01065	0.00918	1.00000
IBD-ebi-a-GCST004131	19	28	3749	3699	50	0.00673	0.00690	0.12587
aFib-ebi-a-GCST006414	1	53	3540	3489	51	1.00000	1.00000	1.00000
SBP-ukb-a-360	0	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	26	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	10	208	13763	13584	179	0.13933	0.14102	1.00000
HTN-panukb	0	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	9	119	10616	10471	145	1.00000	1.00000	1.00000
RA-panukb	21	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	10	122	10154	9976	178	1.00000	1.00000	1.00000
ATH_gtexukb	21	21	3276	3227	49	0.12593	0.12772	0.30000
BMI-panukb	0	53	9215	9114	101	1.00000	1.00000	1.00000
HB-panukb	10	90	8649	8522	127	1.00000	1.00000	1.00000
T2D-panukb	8	14	2460	2423	37	0.04445	0.03957	0.47765
SCZ-ieu-b-5102	7	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	0	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	0	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	1	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	1	4	726	717	9	1.00000	1.00000	1.00000
NS-ukb-a-230	0	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	0	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	10	243	11369	11103	266	1.00000	1.00000	1.00000

Binary, truncated top 1000 genes

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc1000_prob0.5_all.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.5, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.5, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	52	73	5394	5321	73	0.00054	0.00035	1.00000
IBD-ebi-a-GCST004131	71	28	3749	3699	50	0.00100	0.00082	0.44778
aFib-ebi-a-GCST006414	26	53	3540	3489	51	0.00098	0.00081	0.61790
SBP-ukb-a-360	29	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	111	17	2855	2822	31	1.00000	1.00000	0.53280
Height-panukb	49	208	13763	13584	179	0.00478	0.00462	0.37882
HTN-panukb	17	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	44	119	10616	10471	145	0.37442	0.36420	1.00000
RA-panukb	37	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	29	122	10154	9976	178	0.00348	0.00326	0.30765
ATH_gtexukb	105	21	3276	3227	49	0.06320	0.06139	0.57803
BMI-panukb	23	53	9215	9114	101	0.10847	0.09920	1.00000
HB-panukb	77	90	8649	8522	127	0.02120	0.02178	0.17089
T2D-panukb	15	14	2460	2423	37	0.07129	0.07233	0.27451
SCZ-ieu-b-5102	96	15	3547	3511	36	0.01066	0.00979	0.57143
BIP-ieu-b-5110	31	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	9	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	1	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	11	4	726	717	9	0.00080	0.00082	0.07692
NS-ukb-a-230	36	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	27	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	67	243	11369	11103	266	0.04480	0.04144	0.71413

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc500_prob0.6_all.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.6, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.6, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	27	73	5394	5321	73	0.00035	0.00023	1.00000
IBD-ebi-a-GCST004131	45	28	3749	3699	50	0.00279	0.00290	0.04306
aFib-ebi-a-GCST006414	13	53	3540	3489	51	0.00891	0.00740	1.00000
SBP-ukb-a-360	2	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	44	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	31	208	13763	13584	179	0.34322	0.33680	1.00000
HTN-panukb	6	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	21	119	10616	10471	145	1.00000	1.00000	1.00000
RA-panukb	25	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	15	122	10154	9976	178	0.16412	0.16678	0.40667
ATH_gtexukb	44	21	3276	3227	49	0.01918	0.01757	0.57803
BMI-panukb	3	53	9215	9114	101	1.00000	1.00000	1.00000
HB-panukb	14	90	8649	8522	127	1.00000	1.00000	1.00000
T2D-panukb	12	14	2460	2423	37	0.06063	0.05609	0.47765
SCZ-ieu-b-5102	37	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	1	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	4	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	1	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	6	4	726	717	9	0.02717	0.02751	0.30769
NS-ukb-a-230	8	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	22	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	27	243	11369	11103	266	0.43541	0.44302	0.47741

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc1000_prob0.7_all.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.7, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.7, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	25	73	5394	5321	73	0.00378	0.00347	0.61977
IBD-ebi-a-GCST004131	24	28	3749	3699	50	0.14502	0.14033	1.00000
aFib-ebi-a-GCST006414	5	53	3540	3489	51	0.04362	0.04423	1.00000
SBP-ukb-a-360	2	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	59	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	15	208	13763	13584	179	1.00000	1.00000	1.00000
HTN-panukb	8	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	19	119	10616	10471	145	1.00000	1.00000	1.00000
RA-panukb	25	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	14	122	10154	9976	178	0.15407	0.15657	0.40667
ATH_gtexukb	43	21	3276	3227	49	1.00000	1.00000	1.00000
BMI-panukb	7	53	9215	9114	101	0.03383	0.02293	1.00000
HB-panukb	58	90	8649	8522	127	0.16151	0.16369	0.41475
T2D-panukb	7	14	2460	2423	37	0.03899	0.03957	0.27451
SCZ-ieu-b-5102	26	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	7	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	4	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	1	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	3	4	726	717	9	0.01637	0.01657	0.30769
NS-ukb-a-230	12	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	13	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	22	243	11369	11103	266	1.00000	1.00000	0.50004

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_binary/results_y_pred_xgboost_gene_score_bin_trunc1000_prob0.8_all.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (XGboost prediction): 0.8, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (XGboost prediction): 0.8, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	17	73	5394	5321	73	0.00136	0.00117	0.61977
IBD-ebi-a-GCST004131	12	28	3749	3699	50	0.07869	0.07971	0.35897
aFib-ebi-a-GCST006414	0	53	3540	3489	51	1.00000	1.00000	1.00000
SBP-ukb-a-360	0	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	22	17	2855	2822	31	1.00000	1.00000	1.00000
Height-panukb	11	208	13763	13584	179	1.00000	1.00000	1.00000
HTN-panukb	7	42	5465	5400	65	1.00000	1.00000	1.00000
PLT-panukb	13	119	10616	10471	145	1.00000	1.00000	1.00000
RA-panukb	23	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	11	122	10154	9976	178	1.00000	1.00000	1.00000
ATH_gtexukb	27	21	3276	3227	49	1.00000	1.00000	1.00000
BMI-panukb	1	53	9215	9114	101	1.00000	1.00000	1.00000
HB-panukb	37	90	8649	8522	127	1.00000	1.00000	1.00000
T2D-panukb	1	14	2460	2423	37	1.00000	1.00000	1.00000
SCZ-ieu-b-5102	22	15	3547	3511	36	1.00000	1.00000	1.00000
BIP-ieu-b-5110	0	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	1	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	0	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	1	4	726	717	9	1.00000	1.00000	1.00000
NS-ukb-a-230	1	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	1	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	17	243	11369	11103	266	1.00000	1.00000	1.00000

Zscores

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_zscore/results_zfdr0.05_all.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff (FDR  0.05); significant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff (FDR 0.05); significant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	84	73	5394	5321	73	0.00488	0.00328	0.74505
IBD-ebi-a-GCST004131	39	28	3749	3699	50	0.00010	0.00010	0.01435
aFib-ebi-a-GCST006414	8	53	3540	3489	51	0.00016	0.00011	0.61790
SBP-ukb-a-360	4	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	55	17	2855	2822	31	0.21267	0.20525	0.35417
Height-panukb	17188	208	13763	13584	179	0.00084	0.00059	0.32667
HTN-panukb	17	42	5465	5400	65	0.10862	0.10985	0.39252
PLT-panukb	17038	119	10616	10471	145	0.20991	0.21024	0.63206
RA-panukb	24	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	15965	122	10154	9976	178	0.05923	0.05969	0.83051
ATH_gtexukb	45	21	3276	3227	49	0.00218	0.00227	0.02430
BMI-panukb	16080	53	9215	9114	101	0.05838	0.05837	0.33237
HB-panukb	170	90	8649	8522	127	0.01949	0.01726	1.00000
T2D-panukb	10	14	2460	2423	37	0.05527	0.05061	0.47765
SCZ-ieu-b-5102	102	15	3547	3511	36	0.14524	0.13924	1.00000
BIP-ieu-b-5110	0	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	0	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	0	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	3	4	726	717	9	0.01637	0.01657	0.30769
NS-ukb-a-230	2	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	7	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	16214	243	11369	11103	266	0.00178	0.00134	0.84747

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_zscore/results_zfdr0.1_all.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff FDR 0.1, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff FDR 0.1, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	96	73	5394	5321	73	0.00090	0.00057	1.00000
IBD-ebi-a-GCST004131	40	28	3749	3699	50	0.00000	0.00000	0.00466
aFib-ebi-a-GCST006414	14	53	3540	3489	51	0.00003	0.00002	0.36316
SBP-ukb-a-360	5	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	77	17	2855	2822	31	0.22209	0.21008	1.00000
Height-panukb	17228	208	13763	13584	179	0.00082	0.00083	0.32667
HTN-panukb	24	42	5465	5400	65	0.15528	0.15041	1.00000
PLT-panukb	17105	119	10616	10471	145	0.20888	0.20873	0.80531
RA-panukb	24	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	16655	122	10154	9976	178	0.24331	0.24333	1.00000
ATH_gtexukb	48	21	3276	3227	49	0.00266	0.00278	0.02430
BMI-panukb	16569	53	9215	9114	101	0.16964	0.16995	0.49544
HB-panukb	16127	90	8649	8522	127	0.00387	0.00261	0.52833
T2D-panukb	11	14	2460	2423	37	0.06063	0.05609	0.47765
SCZ-ieu-b-5102	141	15	3547	3511	36	0.24179	0.23405	1.00000
BIP-ieu-b-5110	12	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	0	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	0	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	5	4	726	717	9	0.00022	0.00023	0.07692
NS-ukb-a-230	2	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	7	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	17036	243	11369	11103	266	0.01266	0.00972	0.68641

df <- readRDS("/project/xinhe/xsun/multi_group_ctwas/25.multi_group_validation_classfier_0602/results/fisher_xgboost_zscore/results_zfdr0.2_all.RDS")

df %>%
  mutate(across(starts_with("pval_"), 
                ~ifelse(. < 0.05, 
                        cell_spec(sprintf("%.5f", .), color = "red"), 
                        sprintf("%.5f", .)))) %>%
  kable(escape = FALSE, format = "html", 
        caption = "Benchmark gene cutoff FDR 0.2, sigificant ones are highlighted") %>%
  kable_styling(full_width = FALSE)

Benchmark gene cutoff FDR 0.2, sigificant ones are highlighted
trait	num_benchmark_genes	num_gene_08p	num_gene_08m	num_gene_05m	num_gene_0508	pval_08p08m	pval_08p05m	pval_08p0508
LDL-ukb-d-30780_irnt	111	73	5394	5321	73	0.00164	0.00104	0.76457
IBD-ebi-a-GCST004131	46	28	3749	3699	50	0.00001	0.00001	0.00466
aFib-ebi-a-GCST006414	21	53	3540	3489	51	0.00000	0.00000	0.11275
SBP-ukb-a-360	5	35	4909	4858	51	1.00000	1.00000	1.00000
T1D-GCST90014023	114	17	2855	2822	31	0.27645	0.26585	1.00000
Height-panukb	17289	208	13763	13584	179	0.00082	0.00082	0.32667
HTN-panukb	31	42	5465	5400	65	0.19337	0.18917	1.00000
PLT-panukb	17174	119	10616	10471	145	0.26397	0.26439	0.80531
RA-panukb	26	5	332	330	2	1.00000	1.00000	1.00000
RBC-panukb	17288	122	10154	9976	178	0.44575	0.44641	1.00000
ATH_gtexukb	59	21	3276	3227	49	0.00320	0.00334	0.02430
BMI-panukb	16882	53	9215	9114	101	0.16700	0.16695	0.49544
HB-panukb	17123	90	8649	8522	127	0.02295	0.02310	0.73861
T2D-panukb	11	14	2460	2423	37	0.06063	0.05609	0.47765
SCZ-ieu-b-5102	154	15	3547	3511	36	0.26745	0.25699	1.00000
BIP-ieu-b-5110	12	9	2660	2622	38	1.00000	1.00000	1.00000
ASD-ieu-a-1185	0	0	107	103	4	1.00000	1.00000	1.00000
ADHD-ieu-a-1183	0	0	256	249	7	1.00000	1.00000	1.00000
PD-ieu-b-7	5	4	726	717	9	0.00022	0.00023	0.07692
NS-ukb-a-230	3	7	1201	1189	12	1.00000	1.00000	1.00000
MDD-ieu-b-102	14	0	799	796	3	1.00000	1.00000	1.00000
WBC-ieu-b-30	17121	243	11369	11103	266	0.02078	0.01606	0.68641

sessionInfo()

R version 4.2.0 (2022-04-22)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: CentOS Linux 7 (Core)

Matrix products: default
BLAS/LAPACK: /software/openblas-0.3.13-el7-x86_64/lib/libopenblas_haswellp-r0.3.13.so

locale:
[1] C

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

other attached packages:
 [1] forcats_0.5.1    stringr_1.5.1    dplyr_1.1.4      purrr_1.0.2     
 [5] readr_2.1.2      tidyr_1.3.0      tibble_3.2.1     ggplot2_3.5.1   
 [9] tidyverse_1.3.1  kableExtra_1.4.0

loaded via a namespace (and not attached):
 [1] Rcpp_1.0.12       svglite_2.1.0     lubridate_1.8.0   assertthat_0.2.1 
 [5] rprojroot_2.0.3   digest_0.6.29     utf8_1.2.2        R6_2.5.1         
 [9] cellranger_1.1.0  backports_1.4.1   reprex_2.0.1      evaluate_0.15    
[13] httr_1.4.3        pillar_1.9.0      rlang_1.1.2       readxl_1.4.0     
[17] rstudioapi_0.13   jquerylib_0.1.4   rmarkdown_2.25    munsell_0.5.0    
[21] broom_0.8.0       compiler_4.2.0    httpuv_1.6.5      modelr_0.1.8     
[25] xfun_0.41         pkgconfig_2.0.3   systemfonts_1.0.4 htmltools_0.5.2  
[29] tidyselect_1.2.0  workflowr_1.7.0   fansi_1.0.3       viridisLite_0.4.0
[33] crayon_1.5.1      tzdb_0.4.0        dbplyr_2.1.1      withr_2.5.0      
[37] later_1.3.0       grid_4.2.0        jsonlite_1.8.0    gtable_0.3.0     
[41] lifecycle_1.0.4   DBI_1.2.2         git2r_0.30.1      magrittr_2.0.3   
[45] scales_1.3.0      cli_3.6.1         stringi_1.7.6     fs_1.5.2         
[49] promises_1.2.0.1  xml2_1.3.3        bslib_0.3.1       ellipsis_0.3.2   
[53] generics_0.1.2    vctrs_0.6.5       tools_4.2.0       glue_1.6.2       
[57] hms_1.1.1         fastmap_1.1.0     yaml_2.3.5        colorspace_2.0-3 
[61] rvest_1.0.2       knitr_1.39        haven_2.5.0       sass_0.4.1

Weighted pathway-based gene set enrichment

XSun

2025-06-04

Introduction

Model

Benchmarks

Testing genes

Fisher exact test

Pathways

Results

GOBP

Binary, not truncated

Binary, truncated top 500 genes

Binary, truncated top 1000 genes

Zscores

GOBP + GOCC + GOMF + KEGG jointly

Binary, not truncated

Binary, truncated top 500 genes

Binary, truncated top 1000 genes

Zscores